Tag: Generative AI

Oracle Fusion’s New Data Extraction Tool: Getting Your Data Out, Faster and Easier Than Ever

~ K8 Mead ~ Leave a comment

Oracle has introduced a significant enhancement to data extraction in Fusion Cloud Applications with the Read-Optimised Data Store (RODS) and the new Data Extraction Tool. Together, these capabilities provide a modern approach to accessing Fusion data for reporting, analytics and integration workloads, without placing additional demand on live transactional systems.

For organisations looking to feed data warehouses, support operational reporting or enable near real-time integrations, this represents an important shift in how Fusion data can be accessed and consumed.

Traditionally, organisations have relied on a combination of BICC (Business Intelligence Cloud Connector), REST APIs and BI Publisher (BIP) reports to extract data from Fusion. Each approach has advantages, but also limitations. BICC provides reliable bulk extracts, although these are batch-based by design. REST APIs can deliver more current information but often require significant orchestration and development effort. BI Publisher allows flexible data extraction, but queries execute against the same database that supports day-to-day business transactions.

RODS addresses these challenges by separating data extraction from transactional processing. Rather than querying the live application database, extracts run against a dedicated read-optimised environment that is continuously synchronised with Fusion Cloud. This allows large-scale data extraction activities to take place without competing for resources with business operations.

RODS, or Read-Optimised Data Store, is built on Oracle Autonomous AI Lakehouse technology and is designed specifically for high-volume, read-intensive workloads. Fusion data is continuously replicated into RODS using Oracle GoldenGate technology, providing a near real-time representation of transactional data. The result is a platform that supports reporting, analytics and extraction activities without impacting application performance.

Importantly, Oracle is deploying RODS to customers through its Oracle Application Platform technology upgrade programme. There is no separate infrastructure to purchase or manage, and customers can request early enablement through Oracle Support where available.

The Data Extraction Tool provides a modern, Redwood-based interface for creating and managing data extracts.

Available from release 26A onwards, the tool enables users to define extracts through a straightforward configuration process rather than complex development work. To enable the functionality, organisations must activate the feature in Setup Manager under the Manufacturing and Supply Chain Materials Management offering. Oracle Support must also complete the backend activation, and the appropriate security roles must be assigned before users can access the application.

Once enabled, the Data Extraction Tool can be accessed from the Tools menu and is organised around three key areas. Extract Definitions allow users to create and maintain data extraction configurations, selecting the business objects, fields and filters required. Extract Schedules provide a straightforward way to automate the execution of extracts at regular intervals, while Extract Jobs offers visibility of running and completed extractions, making it easy to monitor progress, review results and investigate any issues.

Creating an extract is largely a point-and-click process. Users define an extract name, select an output format (CSV or JSON), choose the required business objects, and specify the fields they wish to include. Filters can be applied to narrow the data set, while column names can be renamed to match downstream integration or reporting requirements.

The tool also supports both Descriptive Flexfields (DFFs) and Extensible Flexfields (EFFs), allowing organisations to include custom attributes alongside standard Fusion data. Extract definitions can be saved for reuse, executed on demand, scheduled for regular execution, or exported between environments.

The Data Extraction Tool supports two execution models:

Synchronous processing is intended for smaller data requests that can be completed immediately. Results are returned within the same transaction, making this suitable for focused data retrieval scenarios. Oracle currently applies a 60-second execution limit.

Asynchronous processing is designed for larger extraction workloads. Users submit a job and receive a job reference that can be used to monitor progress. Once complete, output files are delivered to Oracle Cloud Object Storage or UCM.

Each extract includes metadata detailing the extraction period, row counts and any processing errors, helping administrators validate results and troubleshoot issues.

One of the more interesting additions is Oracle’s Data Extraction Query Transformer Agent. Many organisations have invested heavily in BI Publisher reports and custom SQL queries over the years. The Query Transformer Agent helps modernise these assets by converting SQL into Business Object Query Language (BQL), the format used by the Data Extraction Tool. The agent also validates converted queries against the target schema, helping users identify issues before execution. Oracle has also indicated that future releases will introduce SQL performance assessment capabilities to help organisations identify optimisation opportunities before migration.

RODS coverage continues to grow with each quarterly release. In release 26B, Oracle supports more than 1,100 business objects and has mapped over 600 BICC public view objects. Release 26C is expected to increase coverage to more than 1,650 business objects and nearly 900 mapped BICC view objects, extending support across ERP, HCM, SCM and CX.

Organisations should be aware that business object attribute names may not always match existing ADF View Object or BICC naming conventions. However, output column names can be customised, providing flexibility for downstream integrations and reporting solutions.

Release 26D is set to introduce several important enhancements, including support for custom queries across both synchronous and asynchronous extractions, as well as callback capabilities that enable event-driven processing and notifications.

Oracle is also investing in additional self-service capabilities, including richer extract management, reusable query registration, pre-execution testing and AI-assisted query analysis. As these capabilities mature, the Data Extraction Tool is increasingly positioned as Oracle’s strategic approach to data extraction from Fusion Cloud Applications.

For organisations that currently rely on BI Publisher for integration-focused data extraction, or those finding limitations in existing BICC-based approaches, RODS is worth evaluating. By separating extraction workloads from transactional processing, Oracle has created a more scalable and modern foundation for reporting, analytics and integrations. While coverage is still expanding, the direction is clear: RODS and the Data Extraction Tool are becoming central to Oracle’s long-term strategy for accessing Fusion Cloud data.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

Oracle ERP Cloud Financials 26C

~ K8 Mead ~ Leave a comment

It’s quarterly release time again, and there are some genuinely strong updates in Financials this quarter. I don’t often cover ERP features, but I did last quarter for the same reason and 26C feels similar. As always, more may follow later in the month, but here’s what’s been announced so far and what’s caught my attention.

There are three Fixed Asset Agents in this release. The new Retirement Request Assistant makes it much easier for asset custodians to initiate and track asset retirements. Using a guided, conversational approach, users can submit requests for assets assigned to them or search for others using identifiers such as serial or tag number. They can capture key details like retirement date and reason, then track progress from submission through to completion. This removes the need for emails and manual coordination, giving users a simple self-service route that speeds things up and improves accountability.

Supporting this, the Retirement Assistant helps finance teams review and process those requests. Fixed asset accountants can manage retirements through the same conversational interface, whether dealing with individual assets, multiple assets or file-based uploads. The assistant guides users through the required steps, validates the data, and highlights failed transactions so they can be corrected and resubmitted without starting again. The result is less manual effort, less rework, and quicker, more controlled processing.

The assistant also improves how exceptions are handled. Finance users can review requests, update key details such as retirement dates or proceeds of sale, and post transactions directly within the same experience. Because everything is handled in one place, there is less need to switch between screens or rekey data, which helps reduce errors. Overall, it creates a smoother end-to-end process while supporting stronger governance as volumes grow.

Alongside this, the Fixed Asset Inquiry Assistant offers a much more intuitive way to access asset information. Users can ask questions in natural language to retrieve details across financials, depreciation and distributions, as well as view current period activity. This makes it easier to understand asset movements, validate transactions and respond to audit queries without relying on multiple reports. Taken together, these assistants represent a clear step forward in usability, helping teams reduce effort while improving visibility and control across the asset lifecycle.

The Budget Adjustment Assistant introduces a more straightforward way for budget office users to manage EPM control budgets. Using natural language, users can create and review budget entries, add or reduce budgets, transfer amounts between accounts or periods, and review balances or previously approved entries. The assistant also flags and helps resolve issues at the point of entry, reducing the likelihood of errors and avoiding rework later.

For organisations, this translates into better efficiency and control. Users no longer need to navigate multiple forms or screens, which speeds up processing and reduces effort. At the same time, built-in validation improves data quality before transactions reach the ledger. The result is faster adjustments, fewer issues, and a more streamlined experience for teams managing complex budgets.

The next two enhancements build on the Expenses Agent introduced previously, extending its conversational, touchless approach into more complex scenarios. Cost allocation now allows users to split expenses across multiple cost centres, projects or tasks directly within the agent. Instead of manually distributing costs across lines, users can simply instruct the agent how to allocate amounts. This improves both accuracy and efficiency, ensuring costs are recorded correctly with far less effort.

Another useful addition is the ability to apply cash advances during expense submission. Employees can select one or more available advances, or choose not to apply them and provide a justification where needed. The agent also handles rejected or withdrawn reports by automatically removing applied advances and notifying the user, helping maintain clarity throughout the process.

Together, these updates strengthen the Expenses Agent by reducing manual intervention and improving financial control. Organisations benefit from more accurate allocations, fewer unapplied advances, and better visibility where advances are not used. Employees benefit from a simpler, more guided process that keeps expense reporting moving and reduces delays across the end-to-end lifecycle.

As always, Oracle may introduce additional ERP Agents later in the month. If anything else stands out, I’ll share a follow-up once the full picture is clearer.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

Oracle’s HCM Professional Concierge: What It Is, How It Works, and Why It Matters for HR Teams

~ K8 Mead ~ Leave a comment

Oracle has been steadily building out its AI story in HCM Cloud, but the HCM Professional Concierge is one of the first examples that really feels tangible for HR teams. This is not AI added for the sake of it. It is a set of purpose-built, role-aware conversational agents, built directly into the HCM Redwood experience. For me, it stands out as one of the more considered uses of AI Agents in enterprise HR.

If you work in HR operations or as an HR Business Partner, the scenario will feel familiar. A manager wants to understand where their team sits on compensation ahead of a salary review. They open Employment Info, scroll through individual records, try to piece together performance data from one place, compensation history from another, and absence data from somewhere else. It is not a difficult task, but it is a fragmented one. Before long, ten minutes have passed just getting a basic view.

The HCM Professional Concierge simplifies this by bringing everything into a single conversational experience, embedded wherever the HR user is already working. Instead of navigating between screens, they ask a question. The agent brings together the relevant data, guides the next step, and in some cases can even trigger the action directly from the conversation.

It is worth understanding that this is not a single AI agent working behind the scenes. Oracle has taken a supervisor and sub-agent approach, where a top-level Concierge Supervisor receives the user’s query, interprets the intent, and then routes it to the most appropriate specialist agent.

Within the HR Professional Concierge, those specialist agents each focus on a particular area of HR. For example, the Compensation Advisor brings together key information such as compensation data, compa‑ratios, time since the last salary change, and pay grade details for a manager’s direct reports. The Talent Advisor focuses on performance, helping to summarise ratings and support more informed performance conversations.

Other agents support core HR data and processes. The Employment Details Assistant provides access to employment history, assignment information and worker details, while the Leave and Absence Analyst helps identify and manage absence across a team. There is also support for understanding organisational design through the Workforce Structures agent.

In addition, the Concierge can surface policy and guidance through the Policy sub-agent, review personal worker data where needed, and launch reporting through the Reports sub-agent. For broader, team-level insight, the Team Data Hub helps bring data together to support analysis.

What this means in practice is that the user experiences a single, coherent conversation, even though multiple specialist agents may be working in the background to fulfil the request.

So when a manager asks, “show me the most recent performance rating and time since the last salary change for my direct reports”, the Manager Concierge Supervisor recognises that the query spans both compensation and talent data. It then coordinates across the Compensation Advisor and the Talent Advisor behind the scenes. What comes back is a single, joined-up view, rather than two separate outputs that the manager has to reconcile themselves.

That orchestration across multiple agents is where the real value starts to show. Conversational assistants in enterprise applications are not new in themselves. What is more interesting here is the ability to coordinate specialist agents within a single interaction, carry context across the conversation, and route requests intelligently based on both the topic and the data required.

Oracle has introduced three distinct Concierge experiences, each designed around a specific user group and how they typically work. The HCM Professional Concierge is aimed at HR specialists and HR Business Partners. It sits within the HCM Professional Activity Centre, which has become the central workspace for HR service delivery, and supports the sort of queries an HR analyst would usually run. That includes pulling together workforce data for individuals or manager populations, reviewing compensation and employment history, running reports, looking up policies, and guiding HR actions within the flow of work.

The Manager Concierge is focused on line managers who need quick, straightforward access to information about their teams. It brings together compensation, absence, talent and employment data without the need to navigate into individual worker records. The experience adapts based on both the question being asked and the context of the manager’s team, giving them a practical way to not only view information but also complete common HR tasks directly.

The Worker Concierge, meanwhile, is designed for employees themselves. It brings together support for areas such as leave, payroll, benefits and compensation into a single, consistent experience. Behind the scenes, it routes queries to the relevant specialist agent, whether that relates to absence, benefits, pay, or compensation, so the employee does not need to think about where to go to get the answer.

A simple scenario helps bring this to life. A line manager has been told that budget has been allocated for pay rises and promotions across the organisation. Before making any decisions, she wants a clear view of where her team currently stands. Using the Manager Concierge, she can ask a straightforward question in natural language, such as “how long has it been since my direct reports received a pay rise?” The Compensation Advisor returns the answer in a structured, easy-to-read format. She then follows up with a more specific question, “what is Elaine’s compa-ratio?”, and gets a direct response.

Within the same conversation, she can ask for performance ratings through the Talent Advisor and pull through grade information using the Employment Details Assistant. It all happens in one place, without needing to navigate between screens. Multiple specialist agents are working in the background, but from the manager’s perspective it feels like a single, joined-up interaction.

The HR specialist perspective is just as telling. If someone is working on an Employment Info page for a specific worker, they can open the Concierge panel and ask something like, “what is the salary history for Ravi?” or “where is Ravi located?” The response comes back as structured data pulled directly from HCM, without the need to navigate away or open multiple pages.

One question that comes up consistently when Oracle’s AI features are discussed is around data access and security. It is an important one, and the answer here is reassuring. The HCM Professional Concierge works within the same data and functional security model already applied across the HCM Redwood experience. If an HR specialist does not have permission to view a particular employee’s salary in the core application, they will not be able to access it through the Concierge either. There is no separate access layer being introduced. It simply operates within the role-based controls that are already in place.

For organisations working across multiple geographies, the same principle applies. The agent respects the existing configuration of Redwood pages, including any geography-specific policies and legislative requirements. There is also flexibility to tailor how the agent behaves by refining prompts to reflect your organisation’s terminology or local nuances.

The Concierge also sits within a broader shift in how Oracle is shaping the HR user experience. It is alongside the HCM Professional Activity Centre, which acts as a unified Redwood workspace for HR administration. The Activity Centre brings together a more flexible approach to worker search, with filtering, saved views and personalised results. From there, HR specialists can move straight into transactions from a worker’s profile without switching to a separate area. Common actions are surfaced directly in the interface, including access to areas such as the Recruiting Activity Centre, Mass Assignment Change, Mass Legal Employer Change, Payroll Activity Centre and Attendance Violations, which makes it easier to act on information as soon as it is identified.

The Concierge is always present within the Activity Center, giving HR specialists access to conversational support in the context of the work they are already doing.

It also sits within a much broader direction Oracle is taking with role-based, agent-led HR applications. The HR Specialist Workspace is a good example of where this is heading. It builds on the same foundations, but moves towards a Redwood workspace where multiple specialist agents work together to surface relevant insights more proactively.

In that model, the workspace brings together a view of workforce priorities, potential restructuring impacts, compliance alerts, attrition risk and open HR cases. These are drawn from coordinated agent outputs across areas such as Workforce Management, Talent and Learning. The shift here is subtle but important. The agents are not just responding to questions, they are actively identifying what might need attention and presenting it to the user.

There is also a clear emphasis on governance. Audit trails, controls and human oversight are built into how actions are handed off. Oracle is quite deliberate in positioning this around measurable outcomes, with coordinated agent activity and clear decision points. That creates an important distinction from more autonomous AI models. Here, the agents surface and recommend, but people remain firmly in control of decisions and actions.

From an implementation perspective, the HCM Professional Concierge and its supporting agents are delivered as part of Oracle HCM Cloud Release 26C. There is no need to build these capabilities from the ground up. They are available out of the box, with the ability to adapt behaviour through prompt configuration so that it reflects your organisation’s terminology and ways of working.

As ever, I will keep a close eye on how this develops across the HCM suite and share updates as new capabilities emerge. If you are starting to think about how this fits into your wider HCM AI strategy, or you are planning for a 26C upgrade, now is a sensible point to begin that conversation.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

Oracle Fusion Common Features 26C

~ K8 Mead ~ Leave a comment

It’s my favourite time of the quarter, Oracle has just shared what’s coming in release 26C. I don’t usually cover the Common Features updates, but I did last quarter because there were a few genuinely interesting additions, and 26C feels much the same. As always, more may follow later in the month, but here’s what’s been announced so far and what’s caught my attention.

The first feature is one I’m particularly pleased to see. The HCM Data Security Assistant introduces AI into what has traditionally been one of the more complex parts of HCM. It gives a clearer, more contextual view of how data security is set up across roles, security profiles and user access, helping teams understand why someone can or cannot see certain information or carry out specific actions.

What makes this different is how you interact with it. By asking questions in plain language, security teams can review roles, compare access between users and explore configurations without having to manually work through layers of rules. It can even regenerate access control lists and be enhanced with your own internal documentation, so responses reflect your organisation’s standards as well as Oracle’s underlying model.

The real benefit is the time it saves and the confidence it gives. Instead of piecing things together or raising support requests, teams can investigate issues themselves and resolve them more quickly. This reduces delays for end users waiting on access and helps security teams respond more accurately. It also makes it easier to validate configurations and keep access aligned to policy, which is something many organisations still find challenging.

Spreadsheet data loaders tend to divide opinion. People either rely on them heavily or find them frustrating, usually because of the effort involved in finding the right template and getting the data structured correctly. That’s why I’m glad to see the introduction of the HSDL Advisor in 26C.

This brings a more guided approach to spreadsheet-based data loads. Rather than relying on detailed knowledge of templates, users can upload a CSV file and interact with the advisor using natural language. Behind the scenes, it identifies the correct business object, surfaces the right templates, maps the columns, validates the structure and prepares the file before triggering the load. It also provides clear visibility of progress, with links to monitor outcomes, and can answer questions about templates and configuration along the way.

In practice, this reduces much of the effort and risk associated with HSDL. Users no longer need to understand the technical structure in as much detail, and issues can be identified earlier in the process rather than after a failed load. The support for CSV uploads without needing desktop tools also makes it more accessible. Overall, it’s a more straightforward and reliable way to handle data loads, with fewer errors and less rework.

Within AI Agent Studio, one of the updates that stands out is the ability to manage extended agent interactions through long-running sessions. This allows conversations with AI agents to continue over a longer period, rather than needing to be completed in a single session.

This might sound like a small change, but it makes a real difference in practice. Users can step away, gather additional information or switch tasks, and then return to the same conversation without losing context. For more complex queries, where responses take time or require validation, this creates a much more practical way of working.

It also improves traceability. Having a continuous interaction makes it easier to track decisions and understand how an outcome was reached. For organisations starting to embed AI into everyday processes, this kind of control and continuity is important.

Another key change in 26C is the move away from the AI Configurator to AI Agent Studio as the single environment for managing AI prompts and agents. Where the Configurator focused on editing prompts in isolation, AI Agent Studio brings everything together in one place, allowing teams to design, test and manage both prompts and agents more effectively.

It builds on what’s already there but gives users more control. You can work with a wider range of models, manage variables more easily and test changes before publishing them. From a governance perspective, having one tool helps ensure consistency and reduces the likelihood of changes being made in isolation.

For organisations already using the AI Configurator, there will be some transition effort, as prompts will need to be recreated and validated. However, the ability to copy them into the new environment does help ease that process. In the long run, this feels like a more scalable and manageable way to support AI across HCM.

The final feature I want to highlight is a smaller one, but it will matter to a lot of organisations. The Redwood Appearance Editor has been updated so that brand colours are applied more accurately across Fusion Applications.

Previously, custom colours didn’t always appear exactly as expected, often looking slightly lighter or darker, which made it difficult to maintain a consistent look and feel. With 26C, the primary colour you define will now be applied more precisely across supported areas such as page headers, provided it meets WCAG accessibility standards.

It’s a subtle improvement, but it helps organisations present a more consistent and professional visual identity within the application. Getting branding right is an important part of employee experience, and this makes that easier to achieve while still maintaining accessibility.

As always, Oracle may introduce additional Common Features later in the month. If anything else stands out, I’ll share a follow-up with the highlights once the full picture is clear.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

Oracle HCM Cloud Compensation 26C

~ K8 Mead ~ Leave a comment

I realised recently that I have never actually written a post analysing a Compensation quarterly release. With 26D being the deadline for moving Compensation worksheets to Redwood, release 26C felt like the right time to change that. There is likely to be more to come over the next few weeks, but for now it is worth looking at what has been announced so far.

The feature many people have been waiting for to support that final move to Redwood is now here. The Redwood Workforce Compensation worksheet brings a much cleaner and more intuitive experience for managers, with everything presented in a single, streamlined view. From the landing page, managers can easily access their worksheets and use built-in capabilities such as audit trails, modelling and target application to review and adjust awards with greater confidence. Filtering, search and layout personalisation make it easier to focus on the right employees and the most relevant data. At the same time, guided information panels and simplified alerts help managers understand what needs attention without being overwhelmed. In practice, this reduces the time spent navigating the system and allows managers to focus on making informed compensation decisions.

The design also improves transparency and control. Managers can switch currencies, review how calculations have been derived, and access supporting information such as compensation history, notes and assignment details without leaving the worksheet. Key summary information remains visible while scrolling, so important totals and budgets are always in view. These changes make the process feel more straightforward and help reduce errors and rework, giving organisations greater confidence in the accuracy and consistency of outcomes.

Another interesting addition is the Workforce Compensation Manager Analyst agent, which introduces a more conversational way for managers to interact with their compensation plans. Rather than moving between multiple worksheets and pages, managers can ask questions in natural language and get immediate answers on areas such as budget position, approval status, due dates and manager-level overages. Because the agent works within the context of a specific compensation plan, the responses are relevant and focused, without the need to interpret multiple screens or reports.

From a manager’s perspective, this removes a lot of the friction from the compensation cycle. It cuts down the number of clicks and removes the need to search for information across the system. Managers can quickly sense-check budgets, track progress and identify issues as they arise, all from a single interaction. This supports faster and more confident decision making and helps keep compensation cycles on track.

The enhancements to Redwood Individual Compensation extensibility give organisations far greater control over how compensation changes are entered and managed. Values can now be defaulted not only when creating proposals, but also when correcting or updating them, using dedicated attributes for each scenario. When combined with the ability to apply validation rules to the same fields, this creates a more structured approach to managing individual compensation. These capabilities sit across key processes such as hiring and promotion, as well as within dedicated compensation pages, ensuring consistent behaviour wherever compensation decisions are made.

For users, this reduces manual effort and helps prevent errors before they happen. Defaulting removes the need to repeatedly enter common values, while validation ensures entries meet organisational policies from the outset. At the same time, visibility of who created or updated a record, along with timestamps, strengthens auditability. This makes it easier to track changes and supports a more controlled and reliable process overall.

Alerts within Workforce Compensation have also been simplified. Instead of a wide range of alert types and icons, everything is now grouped into three clear categories: Error, Warning and Information. Existing alert variations have been consolidated, with only genuinely blocking issues presented as errors. This removes much of the visual noise that previously made alerts harder to interpret, and presents information in a more structured way.

For managers, the benefit is immediate. It becomes much easier to distinguish between issues that require action and those that are simply informational. This helps with prioritisation during the compensation process, reduces the risk of missing something important, and supports smoother progression through the cycle.

The final feature is one that has been requested for some time and originates from a customer idea raised on the Ideas Lab. The new Total Compensation Statements Setup OTBI subject area allows organisations to report on how their total compensation statements have been configured. It provides visibility of key elements such as statement definitions, periods, categories and items, along with the relationships between them. With enriched folders and sub-folders, users can explore the structure in more detail and understand how statements are built, without relying on manual documentation or configuration reviews.

This brings greater transparency and control over statement setup. It becomes easier to answer common questions, such as how many items or categories exist, how they are structured, and how they are distributed. This supports better governance and quicker troubleshooting, particularly when reviewing or refining statement designs. By making configuration data more accessible, organisations can maintain consistency in how total rewards are presented and reduce the effort required to manage these statements over time.

As mentioned earlier, Oracle is expected to release additional updates later this month. If anything particularly impactful is introduced, I will share a further update with a more detailed view.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

Oracle HCM Cloud Recruit 26C

~ K8 Mead ~ Leave a comment

With the final deadline for Redwood Recruiting having passed in 26B, the 26C release introduces further innovation, with a strong focus on AI alongside additional Redwood pages to improve the overall user experience. With that in mind, let’s take a look at what’s coming for Recruiting in 26C. As always, Oracle may introduce additional features as the quarter progresses, and if anything particularly notable appears, I’ll share a follow-up update.

The first feature I want to highlight introduces an AI-driven approach to initiating sourcing activity directly from the Recruiting Activity Centre. Oracle has provided a preconfigured agent template designed to sit within recruiting workflows and respond to key requisition activities. Once set up, the agent can automatically trigger when certain activity states are reached, such as when a requisition is awaiting submission, formatting, posting or approval. At that point, the agent can create candidate pools and, if activity slows, generate recruitment campaigns, for example where there have been no recent applications. This all runs in the background once the activity is assigned and the relevant scheduled process is running, removing the need for recruiters to step in manually.

The benefit here is removing friction at the start of the sourcing process and maintaining momentum without relying on manual intervention. Routine but essential tasks such as building candidate pools or launching campaigns are handled automatically and at the right time. This allows recruiters to focus on higher value activities such as engaging with candidates and hiring managers. It also introduces a more consistent approach to sourcing, with the same logic applied across requisitions. Over time, this should lead to faster pipeline creation, fewer delays in attracting candidates and a more proactive hiring approach, even when activity slows.

The next feature, Activity Centres: Automatically Launch AI Agents from Activities, builds on this by making AI agents a standard part of how Activity Centres operate across Recruiting, Sourcing and Interview processes. Organisations can configure published HCM workflow agents to respond automatically when activities are generated, carrying out actions or sending notifications without manual input. By assigning an agent to an activity through a simple configuration and running the scheduled process, the system can identify activities that require action and trigger the relevant agent. As the agent progresses, it updates its status so the system can track whether tasks are in progress, complete or need to be retried.

The value here is in keeping workflows moving without constant user involvement. Activities no longer sit waiting for someone to pick them up or remember the next step. Instead, the system helps move tasks forward and keeps stakeholders informed. This leads to quicker task completion and more consistent execution across recruitment stages, helping to reduce delays in the hiring process. It also improves coordination between recruiters, sourcing teams and interviewers by reducing the reliance on manual follow-ups and emails.

Another key update introduces a more structured and scalable way to support AI-driven content across the external candidate experience. Oracle has moved away from the earlier Prompt Lab approach and standardised on workflow agents that sit behind AI Assist capabilities on career site pages. These preconfigured agents, managed through AI Agent Studio, support a range of scenarios including generating job summaries for search, creating career site content, surfacing relevant assets on job descriptions and providing job fit insights to candidates. As these agents are designed to be reusable and channel agnostic, they can be applied consistently across the candidate journey while still allowing organisations to tailor them.

From a business perspective, this creates a more flexible and modern foundation for candidate engagement. Content generation becomes easier to manage and more consistent, reducing duplication and manual effort. Candidates benefit from richer and more relevant information, helping them better understand roles and suitability before applying. Importantly, this aligns with Oracle’s longer-term direction, giving customers a clearer path forward with a solution that is easier to extend, maintain and evolve as AI capabilities continue to mature.

Smart Search is another enhancement that improves the job search experience in a practical way. Rather than being limited to fixed locations, candidates can now search based on proximity to any location that matters to them, including their current position. The introduction of a search radius provides a more realistic view of available opportunities, helping candidates focus on roles within a reasonable commute. Features such as browser-based location detection also remove friction, making it quicker to find relevant roles.

This is likely to improve both candidate satisfaction and application quality. Candidates are more likely to find roles that genuinely fit their circumstances, while organisations benefit from more relevant applicants. As Smart Search is expected to become the default in a future release, it is worth reviewing your current configuration, particularly the use of fixed versus proximity-based search, to ensure it reflects how your workforce operates.

For organisations using ‘Apply with Indeed’, it is important to plan ahead as this functionality will be discontinued in 27A due to changes in Indeed’s integration model. Transitioning to Direct Apply will help avoid disruption and provides a more seamless and modern candidate experience while ensuring continuity as the legacy functionality is retired.

Turning to Redwood enhancements, the Generate Job Requisition Posting Description Using AI Agent feature brings AI-assisted job description creation directly into the Redwood experience. Using a workflow agent, content can be generated across key sections such as the summary, responsibilities and qualifications. For recruiters and hiring managers, this makes it quicker and easier to create clear and consistent job adverts without starting from scratch.

The shift to an agent-based approach also provides a more robust and future-ready foundation, replacing earlier prompt-based methods while maintaining a familiar experience. For organisations, this means reduced manual effort, greater consistency in how roles are presented and ultimately a stronger candidate experience through clearer, more engaging job descriptions.

Another update, driven by customer feedback through the Ideas Lab, is Bulk Candidate Creation by Uploading Resumes. This allows recruiters to upload multiple CVs at once, with AI extracting key information and automatically populating candidate profiles. Rather than manually entering details for each individual, recruiters can review and refine the extracted information within the candidate record, balancing efficiency with data quality.

The benefit here is a significant reduction in administrative effort and a faster turnaround from receiving CVs to having candidates ready in the system. It also helps teams manage higher volumes more effectively, particularly during peak recruitment periods, while ensuring consistency in how candidate records are created.

Finally, Bulk Actions on the Job Requisitions List makes it easier to manage large volumes of requisitions. Recruiters can now update multiple records in a single action rather than working through each one individually. This includes moving roles through the lifecycle, opening them for sourcing or updating hiring teams. The ability to update the hiring team across multiple requisitions is particularly useful when responsibilities change, removing repetitive manual updates.

For users, this delivers a clear improvement in efficiency and scalability. Bulk actions reduce time spent on administrative tasks and help teams keep pace during busy periods, while asynchronous processing allows larger updates to run smoothly in the background. The result is a more streamlined and consistent way of managing requisitions at scale.

Oracle often introduces additional features as the quarter progresses, so it is worth keeping an eye out for further updates. If anything particularly impactful emerges, I will share a follow-up. In the meantime, if you are planning your 26C adoption or want to explore the updates in more detail, take a look at my latest write-up covering the Core HR enhancements.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

Oracle HCM Cloud Core HR 26C

~ K8 Mead ~ Leave a comment

It’s that time in the quarter again. Oracle has just shared what’s coming in Release 26C, and as you’d expect there is a clear emphasis on AI-driven enhancements, with plenty to take note of. More updates are likely to follow over the coming weeks, but for now it is worth taking a look at what has been announced so far.

The first feature worth calling out is one many HR teams have been waiting for. Ever since the original Person Management page was effectively retired back in 2020, there has not really been a single, equivalent experience to replace it. What Oracle has delivered here is not a direct like-for-like replacement, and it is clear that is not the intention. Instead, this move aligns with the Redwood design approach and rethinks how users interact with person data. The result is a more modern, task-focused experience that brings together key information and actions in a way that feels far more consistent with the rest of the HCM suite.

In practical terms, it changes how users work day to day. Rather than navigating across multiple pages or relying on memory to find the right option, a HR user can open a worker’s record and see relevant insights alongside the actions they are most likely to take, whether that is updating employment details, reviewing assignments, or initiating changes. This reduces the need for constant clicking and context switching, making everyday tasks quicker and more intuitive. For those who have used the system for a long time, there may still be a sense that the original “one stop shop” is missing. However, this is a clear step forward. It signals a shift away from a single static page towards a more guided and contextual experience that better reflects how people actually work today. It may not replicate everything that Person Management once offered, but it goes a long way towards closing a gap that organisations have felt for some time.

The next feature worth highlighting is AI-led, the Positions Management Assistant. This builds on the Positions Assistant introduced in 25D but takes things a step further by broadening what can be done through a single experience. Rather than focusing only on creating or viewing positions, it now brings editing into the same flow, giving HR teams and line managers a more complete way to manage positions. It uses natural language to understand what the user is trying to do and responds with relevant actions and guidance, which fits with the wider move towards more conversational, AI-driven ways of working.

In practice, this simplifies what can often be a fragmented process. A manager can ask to see vacant positions within their team and the assistant will surface them, suggest next steps and provide a direct route to raise a requisition. If a new position is needed, it can guide the user through creating one, reusing existing information where possible to save time and reduce errors. This makes the experience far more intuitive, particularly for occasional users, while also helping to improve consistency and accuracy. The overall result is a smoother, more guided approach that reduces manual effort, speeds up position management activities and helps organisations move more quickly when filling roles.

The next feature builds on something we saw introduced in 25D. The Onboarding Agent has now evolved into a workflow agent, marking another step in Oracle’s move towards more intelligent, guided employee experiences. The new Onboard Assistant takes the earlier self-service capability and turns it into a more interactive, conversational experience for new hires. Rather than working through static checklists or searching for information, users can ask questions in plain language and receive clear, relevant answers tailored to their role, location and organisational policies.

In practice, this makes the onboarding journey feel far more straightforward. A new starter can ask what tasks still need to be completed, check for anything overdue and follow a direct link to take action. The assistant can also surface useful resources, provide reminders and guide users through more complex steps when needed. For anyone unfamiliar with the system, this removes much of the uncertainty that can come with getting started. Overall, it creates a more supported and personalised experience that helps new hires get up to speed more quickly, reduces confusion and ensures that key onboarding activities are completed on time.

There are a couple of notable updates in Document Records, particularly around the use of AI. The first is the Document Records Administration Assistant, which is a clear example of how Oracle is embedding AI into everyday HR administration. In this case, the focus is on simplifying how users retrieve document records. Rather than working through multiple screens and manually applying filters, users can describe what they need in plain language. The assistant interprets the request and submits the appropriate mass download action, removing much of the effort from what has traditionally been quite a manual process.

In practice, this makes a real difference. A HR user could ask to download all passports created in the past month or retrieve payslips generated in the last week, and the assistant will identify the document type, apply the relevant criteria and trigger the correct request. For infrequent users in particular, this removes the need to understand the underlying navigation. The process becomes far quicker and far more straightforward. Overall, it provides a more intuitive way to retrieve document records, reducing admin time, improving accuracy and helping users get to the right outcome first time.

The second update focuses on using AI to extract data from attachments and prefill document record attributes. It is a relatively simple enhancement, but one that addresses a very common pain point in HR administration. When creating document records, users often have to rekey information that already exists in the uploaded file. This feature reduces that effort by using AI to identify and extract key details from the attachment and populate the relevant fields automatically. It aligns closely with the Redwood approach of reducing manual input and making everyday tasks quicker and easier to complete.

In practice, when a user uploads something like a passport or certification, the system can pick up details such as the document number, issuing country and validity dates, and populate these directly into the record. The user can then review and amend the information before saving, rather than starting from scratch. This not only saves time but also reduces the likelihood of manual errors. Over time, capabilities like this can have a noticeable impact on data quality, while also making the process far more efficient for both HR teams and employees managing their own records.

There are a number of upcoming changes around Redwood pages, and the timelines are now starting to feel very close. From 26C, the Redwood person pages will be enabled by default, covering key areas such as Personal Details, Contact Information, Identification Details, Family and Emergency Contacts, Additional Person Information and Person Identifiers for External Applications. For many organisations, this is the point where the move to Redwood becomes unavoidable for core HR data.

This continues in 26D, with areas such as Jobs, Locations, Departments, Enterprise HCM Information, Grade Ladders and all Employment pages, including actions like Add Assignment and Employment Information, also switching to Redwood by default. The same pattern carries into 27A, where processes such as Resign from Employment, Mass Assignment Change and Terminate Employment will be automatically enabled.

If you have not already moved to these Redwood pages, now is the time to start planning. Leaving it until they are switched on by default means losing control over when the change happens and removes the option to step back if needed. Moving earlier gives you the opportunity to test properly, prepare your users and resolve any issues before the transition becomes mandatory.

As mentioned earlier, Oracle is expected to release additional Core HR updates later this month. If anything stands out as particularly impactful, I will share a further update with a more detailed view. In the meantime, keep an eye out for the upcoming posts in this series where we will explore other areas of Fusion as part of Release 26C. If you are reviewing your own roadmap or considering how these changes might affect your organisation, now is a good time to start the conversation.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

A Day with Oracle: AI Success Navigator and Guided Learning Partner Enablement

~ K8 Mead ~ Leave a comment

Today I had the opportunity to attend and present at a partner enablement event hosted by the Oracle AI Success Navigator product team, focused on how partners like Version 1 can best use Oracle’s tooling to bring genuine, measurable value to our customers. The session brought together presentations, product demos, hands-on labs, and open discussion, covering Oracle Cloud Success Navigator and Oracle Guided Learning (OGL). It was a useful day, and I wanted to share some of the key takeaways while they’re fresh.

If you haven’t come across Cloud Success Navigator yet, it’s Oracle’s digital engagement platform, provided free to Oracle Fusion Cloud customers, designed to help organisations design, implement, and accelerate their cloud and AI roadmaps. It sits at the centre of Oracle’s broader AI Factory offering, which Oracle launched as a bundled set of partner and customer services aimed at speeding up AI adoption.

At its core, Cloud Success Navigator gives customers a single place to discover new features, plan adoption, track key milestones, and access Oracle Modern Best Practice (OMBP) guidance. The sunburst visualisation is particularly useful: it surfaces relevant features based on your production profile, so your team isn’t wading through capabilities that don’t apply to your configuration. You can tag features across Now, Next, and Later columns, which gives a clean, structured view of your innovation roadmap.

A significant addition to the platform is AI Assist, which was made generally available in late 2025. AI Assist is a generative AI-enabled assistant embedded throughout Navigator. It goes beyond a standard chatbot: it provides tailored recommendations, surfaces relevant documentation, highlights release roadmap changes based on your context, and flags project milestone risks. For partners, the practical implication is that our customers now have a self-service layer of intelligent guidance that can accelerate feature discovery and planning without always needing to raise a support request or wait for a consultant touchpoint.

How should Partners be using Success Navigator? This was, for me, the most valuable part of the day. The Oracle product team was clear that Navigator is not just a tool for customers to log into independently. The expectation is that partners should be actively bringing Navigator into their delivery model, whether that’s during implementation, post go-live optimisation, or ongoing managed service.

In practice, that means a few things. During implementation, your partner should be walking you through Navigator as part of onboarding, not treating it as a nice-to-have that gets mentioned at the end of a project. Feature planning sessions are more productive when they’re anchored in Navigator’s release data and OMBP content, rather than relying on spreadsheets or static documentation that goes out of date.

Post go-live, Navigator becomes a continuous value tool. The AI Assist agents can help customer teams stay ahead of quarterly release content, plan for Redwood migration milestones, and identify AI features that fit their production profile. Partners who are actively guiding their customers through this ensure their customers are in a much stronger position than those who are leaving customers to self-serve without direction.

One thing to note: Oracle has indicated that the platform continues to evolve, with enhancements planned around streamlined account management for customers with multiple accounts and improved programme management views. It’s worth keeping an eye on the in-application release announcements for Navigator itself.

The second major focus of the day was Oracle Guided Learning (OGL), Oracle’s digital adoption platform (DAP) built natively for Oracle Cloud applications. OGL delivers in-application guidance, directly overlaid onto the Oracle Fusion interface, so users get real-time, contextual help without having to leave the system or refer to separate documentation. The core capabilities OGL brings to a customer environment are worth spelling out clearly, because I still encounter organisations that underestimate what the platform can do.

Process guides provide step-by-step walkthroughs for complex transactions, walking a user through the exact steps required to complete a task within the application. Smart tips and beacons offer contextual pop-up hints and visual cues at key points in the UI. The Help Panel gives users access to self-service guidance and documentation from within the application. In-app messaging allows administrators to send announcements, policy updates, and maintenance communications directly to users as they work, rather than relying on email campaigns that often go unread. Analytics then close the loop: OGL captures how users are engaging with content, where they’re dropping off, and which features or processes need additional guidance investment.

What’s particularly relevant for customers right now is the AI integration within OGL. The OGL 26A release introduced generative AI capabilities into the content authoring experience: content developers can use an AI assistant within the Full Editor to generate and rephrase step text for process guides, smart tips, beacons, and messages. This significantly reduces the time needed to build and maintain a library of guides, which has historically been a barrier to adoption on smaller or resource-constrained engagements.

OGL also extends beyond Oracle applications. It can be deployed across third-party applications including Salesforce, ServiceNow, Microsoft SharePoint, and others, which is useful context for customers running a mixed application estate.

A thread running through both topics today was change management, and it’s one that I think partners sometimes treat as a soft add-on rather than a structural part of delivery. The reality is that both Navigator and OGL exist precisely because technology adoption is a change management problem as much as a technical one.

Navigator gives you the roadmap visibility and planning structure to keep customers engaged with what’s coming and why it matters. OGL gives you the in-application mechanism to reinforce new behaviours, communicate changes, and support users at the moment of need. Used together, they cover a significant portion of the adoption lifecycle: from feature discovery and prioritisation, through to in-system guidance and analytics-driven optimisation.

The enablement message from Oracle today was straightforward: partners who embed these tools into their delivery model are better placed to demonstrate continuous value to customers. Customers who have a structured adoption programme, supported by Navigator and OGL, tend to see higher feature utilisation and lower support overhead than those who treat go-live as the end of the engagement.

It was a practical and well-structured day. The Oracle AI Success Navigator product team clearly has a strong vision for how the platform should be used within the partner ecosystem, and the investment Oracle has made in AI Assist and the broader AI Factory infrastructure is evident. For those of us working in Oracle Fusion Cloud implementations and managed services, the message is clear: these tools are available, they’re free as part of the Oracle subscription, and using them well is increasingly a differentiator in how we position value to our customers.

If you’re currently working on an Oracle Fusion Cloud engagement and you haven’t had a detailed look at what Cloud Success Navigator and OGL can offer, now is a good time to start that conversation.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

Going Deeper with Oracle AI Agent Studio: Connecting, Triggering, and Building with Confidence – Part 4

~ K8 Mead ~ Leave a comment

Over the last three blogs, I’ve explored how AI Agent Studio connects to the wider enterprise, how agents are triggered and interacted with, and how workflows are designed to be reliable and production‑ready. In this final part of the series, I want to pull those threads together and focus on the capabilities that help agents scale safely and operate with confidence over time. This is where governance, control and operational discipline really come into play, and where the newer 26A and 26B features start to show how Oracle is shaping AI Agent Studio for long‑term, enterprise use rather than short‑lived experimentation.

Choosing the right document or memory node is an area where I see a lot of confusion in conversations with clients, so it is worth being very clear about what each one is designed to do. The Document Processor node is intended for runtime documents, attachments that arrive as part of a specific workflow execution, such as a supplier quote received by email, an invoice uploaded through chat, or a UCM attachment linked to a Fusion business object. Its job is to retrieve the file, extract the text, and pass that content on to the next node in the workflow. It is not designed for querying a stable or long‑lived corpus of documents, such as policy or reference material that you want to reuse and search repeatedly over time.

The RAG Document Tool node is designed for exactly that stable, reusable collection of information. You curate a set of documents within an Oracle AI Agent Studio Document Tool, move them through the lifecycle from Ready to Publish to Published, and the RAG node then performs semantic retrieval against that content to ground downstream LLM reasoning in your own policies, playbooks or manuals. To get the best results, it is important to use specific queries with clear discriminators such as module, process area, country or version, which helps improve retrieval precision. It is also good practice to include an explicit “no results” fallback path in your workflow, rather than allowing the LLM to guess when retrieval confidence is low.

The Vector DB Reader and Writer nodes serve a different purpose again, providing durable semantic memory that persists across workflow runs. They are best used to store normalised, reusable knowledge units such as validated resolution summaries, previous exception details, or extracted entity representations. Entries should be kept short and semantically focused, enriched with meaningful metadata to support filtering, and assigned stable document IDs to avoid duplicates. Raw PII or permission‑restricted data should never be stored without a deliberate access control design. When reading from the vector store, metadata filters should always be applied, and low‑confidence matches should be treated the same as no result at all, routing the workflow to a deterministic fallback rather than continuing on uncertain ground.

One theme that came through strongly in the partner training sessions, and one I think represents genuinely good discipline, is treating Workflow Agent testing as a first‑class concern rather than something bolted on at the end. Oracle’s evaluation framework for Workflow Agents, often referred to as Workflow Evals, is based on supplying structured JSON test inputs and asserting expected outputs. These evaluations are intended to be run as a regression suite whenever you change a prompt, adjust a node configuration, swap a tool, or update a policy, helping you catch unintended side effects early and keep agent behaviour stable as it evolves.

A good starting point is to define around five core paths through the workflow: the happy path, two or three of the most common exception scenarios, and at least one case that deals with missing or poor‑quality input data. From there, you should be tracking things like overall pass rate, branch accuracy, schema validity, and retry or escalation behaviour. The aim is not simply to prove that the workflow reaches an end state, but to make sure it routes correctly and predictably under every condition that genuinely matters in production.

For anyone building more complex workflows, the full context variable reference is well worth bookmarking. In practice, a small set of variables tends to do a lot of the heavy lifting, such as $context.$nodes.<nodecode>.$status to check whether a preceding node succeeded or failed, and $context.$nodes.<human_node_code>.$actionPerformed to capture whether a Human Approval step resulted in APPROVE, REJECT or REQUEST_CHANGES. You can also use $context.$nodes.<human_node_code>.$feedbackReceived to pick up any comments provided by the approver, and $context.$workflow.$traceId to generate idempotency keys or include trace references in error notifications. For conversational workflows, $context.$system.$chatHistory is particularly useful, as it exposes the full session history and allows the agent to reason about what has already been discussed.

The 26A roadmap also includes several upcoming capabilities that will significantly extend what is possible in the near term. Support for the Model Context Protocol, or MCP, means Workflow Agents will be able to invoke tools exposed by MCP servers, broadening the integration landscape well beyond traditional REST APIs. The Agent Studio Help Assistant, an AI‑driven guide embedded directly within the studio, should also make agent design far more accessible, particularly for practitioners who are new to the tooling. Alongside this, multi‑modal enhancements, including end‑user Q&A over images and documents uploaded in chat and semantic search across non‑text assets, open up an entirely new set of document understanding and reasoning use cases.

Looking a little further ahead, the roadmap includes capabilities such as breakpoint‑style debugging, automated prompt engineering, multi‑user development environments, and a Bring Your Own LLM option, alongside additional interaction channels including WhatsApp, SMS and telephony. Taken together, these signal a sustained level of investment in the platform and a clear focus on making AI Agent Studio more powerful, more accessible, and more suitable for enterprise‑scale use. The overall direction is a positive one, and it is clear that Oracle is building towards a mature, long‑term agent platform rather than a short‑term experiment.

The partner training sessions that informed this post covered a lot of practical ground, and I genuinely believe they will save teams a significant amount of time as they start building in earnest. If you are already exploring AI Agent Studio and would like to talk through any of these patterns in more detail, I would be very happy to continue the conversation. And if you have not yet read the earlier posts in this series, it is worth starting at the beginning with the overview of how Workflow Agents are structured, which sets the context for everything covered here.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines

Going Deeper with Oracle AI Agent Studio: Connecting, Triggering, and Building with Confidence – Part 3

~ K8 Mead ~ 1 Comment

In the first two blogs, I looked at how AI Agent Studio connects to the wider enterprise landscape and how agents are triggered and engaged, whether by systems, schedules or users. In this third part, I want to step back slightly and focus on what happens inside the agent itself, specifically how workflows are structured, how context is managed, and how you start designing for reliability rather than experimentation. This is the point where agent design shifts from “can we make it work?” to “can we trust it to run consistently in production?”, and the 26A capabilities give you far more control here than many people realise. To check out the previous blog, please click here.

The Wait node, which is being introduced as part of the 26B release, addresses a long‑standing gap in workflow design, where there was no clean way for a workflow to pause and resume later without either completing immediately or blocking indefinitely. When a Wait node is reached, the workflow moves straight into a Waiting state and pauses execution for a configured period of time, up to a maximum of 60 minutes. Once that wait period expires, the workflow can optionally loop back to an earlier point before continuing, allowing it to re‑evaluate conditions or check for updates. This looping behaviour is controlled through two simple settings: the Loop Back Node, which defines where execution returns to, and Maximum Iterations, which limits how many times the workflow can loop before it continues forward regardless.

In practice, this enables a clean polling pattern that is otherwise difficult to model. For example, imagine a workflow that creates a receipt request in Fusion and then needs to confirm that the receipt has been posted before it can move on. By using a Wait node configured for five minutes and looping back to a Business Object read node up to ten times, the workflow effectively gives itself a 50‑minute window to detect the receipt posting automatically before either continuing or escalating. During each wait cycle, the node outputs ORA_USER_INPUT_REQUIRED, and once all iterations are exhausted it returns WAIT_TIME_EXPIRED_AND_MAX_ITERATIONS_REACHED, both of which can be evaluated in downstream If Condition nodes to route the flow appropriately.

The Code node is one of the most powerful building blocks in a Workflow Agent, and also one of the most commonly underestimated. It executes JavaScript and returns a single value, whether that is an array, boolean, number, object or string. Its real value lies in handling the deterministic work that you should never push into an LLM node, such as data normalisation, threshold calculations, schema validation, array filtering and payload shaping. Used well, it provides a clean separation between predictable logic and probabilistic reasoning, which is a key ingredient in building workflows that behave consistently and are easier to trust in production.

There are a few important constraints to be aware of when designing logic for the Code node. Execution is limited to five seconds, with an upper limit of 100,000 statement executions, and functions cannot be defined within the code, which means recursion is not supported. Most built‑in JavaScript methods are available, but there is no external access, so no REST calls, file system operations, console logging or library imports. The code can read from $context, $currentItem and $currentItemIndex, but it cannot modify the $context object directly. Instead, it simply returns a value, and that returned output is the sole result of the node.

Some of the most effective patterns I’ve seen make particularly good use of the Code node for this kind of deterministic work. Common examples include normalising inconsistent date strings and currency values into canonical formats before passing them to a Business Object write node, or calculating variance percentages for three‑way match validation so that an If Condition node receives a simple boolean rather than needing to express complex arithmetic. Other strong patterns include generating idempotency keys using a combination of $context.$workflow.$traceId and object identifiers to prevent duplicate writes during retries, and filtering arrays returned from Business Object reads so that only active or primary records are passed into a For Loop for further processing.

For workflows that are triggered through the AI chat interface, 26A also introduced support for file uploads during conversations with an agent, allowing users to attach up to five files with a combined size of 50 MB. A wide range of formats is supported, including PDF, DOCX, XLSX, PPTX, PNG, JPEG, HTML, Markdown, JSON, XML, CSV and ZIP. To work with these attachments inside a Workflow Agent, 26A required the delivered MultiFileProcessor tool to be added to an agent and that agent then included within the main workflow. This capability significantly expands what chat‑driven workflows can handle, particularly when dealing with documents, structured data and supporting evidence provided directly by the user. In 26B, this has been simplified significantly. Rather than introducing a separate agent, you can now add a Tool node directly into your Workflow Agent and select Chat Attachments Reader as the tool type. This keeps the workflow much cleaner and removes an unnecessary orchestration step. The tool reads the files uploaded in the current chat session and exposes the extracted content directly to downstream nodes, making it easier to act on user‑provided documents without additional plumbing or indirection.

Support is also in place for third‑party file storage, allowing users to upload files directly from Google Drive, Dropbox or Microsoft OneDrive, provided those credentials are configured under the Chat Experience tab in Credentials. Enabling this involves registering an OAuth application with the relevant provider, obtaining the client credentials, configuring the account in Credentials, and then switching on the option to allow users to upload files from connected cloud storage accounts on the agent’s Chat Experience tab. Once configured, this gives users a seamless way to bring external documents into agent‑driven workflows without needing to download and re‑upload files manually.

This third blog has focused on what really makes Workflow Agents robust in practice, from pausing and polling patterns, through deterministic logic in Code nodes, to handling documents and attachments cleanly inside workflows. These are the building blocks that move agents beyond experimentation and into something you can rely on day to day. In the final post in this four‑part series, I’ll bring everything together and look at the remaining 26A and 26B capabilities that round out the platform, focusing on how they support governance, scale and long‑term operational confidence when running AI agents in production.

Please note all screenshots are the property of Oracle and are used according to their Copyright Guidelines