If you have been anywhere near an Infor conversation this year, you would have heard the word ‘agentic’ more times than you can count. Infor customers are moving from asking what AI can tell them to asking what AI can do for them, and that shift is changing how ERP teams think about their roadmaps.
The good news is that there is no single ‘correct’ way to get there. Whether you are just starting to explore agentic AI or already have a use case in mind, there are three realistic routes into building agents for your Infor system, and which one fits depends on how much you want to build versus adopt and how specific your process is. Here is how we would frame it.
Agentic AI adoption is picking up across the Infor customer base
Across manufacturing, distribution, food and beverage, and fashion the pattern looks similar. Infor customers spent the last few years getting their data foundations right. Now the conversation has moved to the gap between knowing and doing, the hours that pass between spotting an issue and actually resolving it across every system it touches.
That is exactly the gap agentic AI is built to close. Instead of a person reading an alert and manually working through five systems to fix it, an agent can carry that decision through the workflow itself, flagging what needs a human eye and acting on what does not. For ERP consultants and business users who already spend their days on CloudSuite, that is a meaningful change to how work gets done.
Route 1: Build natively with Infor Velocity Suite
Infor’s Velocity Suite is built around the idea of the agentic enterprise, an operating model in which agents do not just surface insights but turn them into action, inside your existing workflows, with governance built in from the start.
Infor frames this around four pillars: agents with real industry context, experiences that adapt to the role and the work at hand, orchestration that lets agents run end-to-end without waiting on manual handoffs, and governed velocity, meaning trusted data and compliance are designed in rather than bolted on later.
For Infor customers, this matters because it means agentic AI is not an add-on you have to bring in from outside. It is native to the platform you are already running on, which is a real advantage if governance and long-term platform alignment are high on your list.
Our Order Capture Agent is a good example of what an agent built for Velocity Suite looks like in practice. It reads incoming purchase orders from email or PDF, validates them against business rules and buying history, and creates clean orders directly in Infor CloudSuite, using Infor Velocity Suite’s RPA, document processing, and GenAI capabilities rather than a separate bolt-on system. Exceptions still go to the order desk for a human check, but the orders that are ready to process no longer need someone to key them in first. It is a clear illustration of how building natively on Velocity Suite means extending the value of the platform you already have, rather than layering something new on top of it.
Route 2: Start with proven base agents
The second route is to use agents built for Infor. Rather than starting from a blank page, Fortude’s base agents give ERP consultants and business users conversational, natural-language access to core M3 modules on day one, replacing the manual lookups that eat into everyone’s time.
Fortude has built a set of these Infor AI Base Agents which you can access here. But to give you a feel for what this route looks like in practice, here are a few of the agents:
- The Sales Agent gives sales reps and managers instant visibility into open orders, customer accounts, and pricing, and flags at-risk orders before they become a problem.
- The Inventory Agent gives planners and warehouse teams a clear read on stock levels and replenishment needs, and flags stockout risk with recommended reorder quantities.
- The Financial Planning and Analysis Agent works the same way for finance teams, explaining budget variance and drafting forecasts in plain language instead of a spreadsheet full of formulas.
The appeal of this route is speed. These agents are purpose-built for Infor, so you are not starting from zero; you are adopting something that already understands your data and modules and can be customized based on your business requirements.
Route 3: Customize agents around your own processes
Not every business runs the same process the same way, and sometimes the answer is not a native build or an off-the-shelf agent, but a version of the base agent (discussed above) shaped around how your business actually works.
This is where the architecture behind our base agents earns its keep. They are not built as rigid, single-purpose tools. The underlying architecture is designed so that the logic, the data sources, and the workflow triggers can be adjusted to match a specific business process, without starting the build over from scratch. For instance, a Procurement Agent built for one customer’s approval hierarchy and supplier scorecards can be reconfigured for another customer’s contract terms and sign-off chain, using the same foundation.
For Infor customers with a process that does not quite match the standard agent capabilities, whether that is an unusual approval flow, a specific compliance step, or a data source the base agent was not built to touch, this route means you are not choosing between buy and build from nothing. You are extending something proven.
Governance has to run underneath all three routes
Whichever route you choose, introducing agents into an ERP environment raises the same set of questions. Who has access to the agent’s source, its skill files, and its plugins, and how do you stop business-sensitive data or company IP leaking through them? How do you avoid shadow IT, where teams quietly deploy their own agents without any structure around cost, security, or maintainability? How do you keep experimentation alive without losing control of what is actually running in production?
The answer is not to slow innovation down. It is a single plane of control, one interface where model access, quotas, and usage can be monitored, where onboarding a new agent or skill file follows a defined process rather than an ad hoc one, and where issues can be identified and resolved before they become bigger problems. This is exactly what Infor’s own framing of governed velocity gets at, and it holds whether you are building natively, adopting a base agent, or customizing one for your own process. It is the thread that needs to run through all three routes from the start.
What’s next on your agentic journey?
Native build, base agent, or customized base agent are not competing options so much as three doors into the same outcome: ERP consultants and business users getting agents that act on their behalf inside workflows they already trust. Which door you walk through depends on how much you want to build versus adopt, and how closely a process needs to match the way your business actually runs.
Whichever route fits, that is the conversation worth having before you build.
Want to talk through which route makes sense for your Infor environment? Get in touch with our team.
