Do you have a question? Want to learn more about our products and solutions, the latest career opportunities, or our events? We're here to help. Get in touch with us.
With AI agents now available across the Dynamics 365 and Power Platform suite, the opportunity for organisations running Dynamics 365 is considerable. But for many, the path from understanding what's possible to implementing agents in production remains unclear.
We spoke to Ralf Wittgen, Customer Success Manager within Datacom's AI Business Solutions team, about what he's seeing in the market, where agents are already making a difference and why understanding your current environment is the essential first step.
What are you seeing in the market right now when it comes to D365 environments and AI adoption?
Adoption is the big theme right now. A lot of organisations have been talking about AI and dabbling in specific areas, but most haven't scaled it out to make AI a core part of how they operate. There are a few larger organisations that have moved things into production, but the majority are still asking: do I trust AI? Do I have the right guardrails? Are my people skilled up?
What we're hearing consistently is that teams understand tools like ChatGPT, but in their day-to-day jobs they either don't have the license to use AI internally, don't know how to apply it, or aren't sure where it can save them time or provide value. That's why we've been running hands-on Copilot enablement sessions, which are helping people break through that barrier of understanding and get hands-on with the tooling in the context of their own roles.
It comes down to two things: skilling and governance. The organisations that are further along have invested in educating their people first, so they understand the foundations before making bigger leaps. And then they've put governance structures in place, including clear guardrails and security principles, before opening things up to the wider organisation.
Without that foundation, you end up with the same pattern we've seen before with Access databases, then SharePoint libraries, then Power Apps. People create things, and suddenly you've got a sprawl of ungoverned apps. Agents are the same. If you give people the ability to create them without the right controls, it gets messy quickly.
Microsoft has been releasing agents that run on top of the D365 platform. In sales, there's a lead qualification agent that can scrape information, understand and surface leads based on your criteria, and triage them to your team; cutting down the time people spend manually managing that process.
In customer service, agents can scan closed cases and automatically generate knowledge articles about how issues were resolved. So next time a similar problem comes in, the answer is already available.
For financial services specifically, we’re seeing a focus on compliance and risk management as key value drivers. Agents can monitor conversations with customers – whether it's about home loans, credit cards, or wealth management – and flag where compliance steps haven't been followed. If someone hasn't completed a required process within a set timeframe, the agent will alert them or escalate to their manager. It's helping organisations stay on top of obligations like the Credit Contracts and Consumer Finance Act without relying solely on manual checks.
The quickest wins come from repetitive tasks that humans are doing on a daily, weekly, or monthly basis – things that can be handled by an agent instead. Beyond that, it's about looking at processes that are disconnected or manual and automating them with AI.
We're also seeing government and public sector organisations looking at this. When you think about responding to citizen requests or processing Official Information Act requests, there's no reason AI shouldn't be able to scan and surface those answers significantly faster.
Data is the big one. If your data is untidy or your structures aren't correct, you're going to have problems when you layer AI on top. It's a bit like error carried forward in maths. If the foundations are wrong, everything built on top inherits those issues.
We worked with a non-profit customer who wanted to move to a contact centre powered by AI, but before they could get there, their base CRM system needed to be in a good state. Relationships in the data were wrong, structures were incorrect. If you bring an AI-powered agent into that environment, it's going to surface the wrong information to customers and that creates a poor experience very quickly.
Most D365 environments have been built up over years by different teams and different partners, and nobody has a complete picture of what's in there anymore. It’s something we see a lot, which is why our team built Decode – a tool that uses AI to reverse-engineer a Dynamics 365 CRM system and produce as-built documentation. It scans through all the system components, so you get a clear picture of what you have, not what you think you have. A process that used to take weeks or months manually now takes a fraction of that time. Our early results show roughly between 60 to 80 percent time reduction.
Get your people comfortable with the tooling, understand the boundaries, and learn the principles of prompting. This builds confidence before you make bigger leaps.
Put guardrails in place, set up a centre of excellence, and define your security principles. Agent 365 can act as your control plane, giving you visibility and control over what's being created across the organisation.
If you're unsure about the state of your current D365 environment, get an as-built picture before layering anything new on top. Decode can do this in a fraction of the time it would take manually, scanning your system and documenting what's there so decisions are based on facts, not assumptions or out of date documentation.
Once the foundation is in place, start identifying repetitive or manual processes that agents could handle, and move them into proof of concept. These can run in parallel with your governance work.
With education, governance and a clear understanding of your environment in place, you can open things up more broadly knowing that AI is being adopted on solid foundations rather than carrying risk forward.
From Dynamics 365 and Power Platform through to Microsoft Copilot, successful AI adoption starts with the right foundations. Discover how Datacom and Microsoft help organisations modernise business platforms, strengthen governance and unlock value from AI with confidence.
