Getting Agent-Ready: How Businesses Can Turn AI Into Practical Value
A conversation with James Oleink, Partner Director of Product Management for Microsoft Agent 365 and Microsoft Dataverse
Building an artificial intelligence (AI) agent is becoming easier. However, the real challenge is knowing where to use one, how to connect it to business systems, and how to manage it as adoption grows. For organizations using Microsoft business applications, success depends on putting AI to work in ways that improve productivity and deliver business value.
In a recent episode of Backstage Pass, my monthly podcast series exploring developments in Microsoft business technology, I sat down with James Oleinik, Partner Director of Product Management for Microsoft Agent 365 and Microsoft Dataverse, to discuss what it means to become agent-ready. Drawing on his experience across Microsoft’s Power Platform, Dataverse, Dynamics, and agent initiatives, James shared practical insights on moving beyond AI experimentation.
Our conversation covered choosing the right AI tools, establishing governance, and connecting business data to help AI automate work without adding administrative burden. The following Q&A explores these topics and the practical steps you can take to make AI a more effective part of everyday operations.
Want to hear the full conversation? Listen to the Backstage Pass podcast with James Oleinik here.
Let's start by introducing our guest, James Oleinik. James is Partner Director of Product Management for Microsoft Agent 365 and Microsoft Dataverse at Microsoft. With 16 years at Microsoft spanning the Power Platform, Dataverse, and Dynamics, he brings a broad perspective on the infrastructure, data, and governance needed to support the next generation of AI-powered business applications and agents.
Q: How can businesses move beyond AI experimentation and start seeing real productivity gains?
James: It’s easy to think of AI as a better search engine for your business information. And while that’s certainly useful, I think it undersells the opportunity. The real shift is one of empowerment. The people closest to the work are often the ones best positioned to identify opportunities to improve and automate it.
That requires a different mindset. The tools are evolving quickly, so businesses need to be comfortable with change, stay curious, and keep experimenting. On my own product team, for example, we were experimenting with agents to automate parts of our processes, such as reviewing feedback from the community. A year later, we’re automating more of that work directly in tools like Copilot Cowork. Even the way we work has changed dramatically in that time.
What made that possible was giving people room to experiment rather than responding to AI with fear or locking everything down. If you give employees clear pathways to try these tools, you may be surprised by what they can accomplish. Start by empowering the people who understand your business and giving them the tools to rethink how work gets done. They’ll often know where to begin.
It does take time and experimentation. There may be some bumps along the way, but once people have that breakthrough and see what’s possible, they can start finding ways to get significantly more value from AI. We’re seeing that at Microsoft, too. Some users are achieving dramatically greater productivity because they’ve figured out how to take their use of AI to the next level. The challenge is helping more people reach that point.
➜ Empowering employees to experiment is a good starting point. But with so many AI tools available, how do you know which one is right for a particular task?
Q: When should businesses use Copilot Chat, Copilot Cowork, or an AI agent?
James: I think it comes down to the type of task you’re trying to accomplish. Each tool has a place in the workflow, and the right choice depends on whether you need help thinking through something, want to delegate a task, or need to make work part of a team’s ongoing business process.
I use Copilot Chat to research, summarize, draft, and think through questions. It’s particularly useful when I’m exploring an idea or working through a problem. Copilot Cowork is more about handing off a task I want completed. That task might take time or involve multiple systems, but I’m still involved and can oversee the work. Agents serve a different purpose. They’re best suited to work that needs to happen consistently across a team as part of a broader business process. The distinction is moving from helping me get something done to helping us work more effectively as a team.
At-a-glance Tool Distinction:
- Copilot Chat: Researching, summarizing, drafting, and thinking through questions
- Copilot Cowork: Delegating tasks that may involve multiple steps or systems, while retaining oversight.
- AI Agents: Supporting repeatable work that becomes part of a team’s business process
There’s a natural progression here. You might start by using Chat to explore a question, then delegate a task to Cowork. When that work becomes a process a team needs to follow consistently, an agent may be the better fit.
➜ As employees begin delegating more work and teams build agents of their own, experimentation can quickly turn into a larger operational challenge. Businesses need to know what has been created, what it can access, and who is responsible for it.
Q: Why do businesses need governance and management for AI agents?
James: As AI agents become easier to create, it becomes more important to understand and manage how they’re being used.
Microsoft Agent 365 is designed to serve as the control plane for your agents. It gives you visibility into what agents exist, who owns them, what systems they can access, and what they’re doing. We’re also working to help you understand the value those agents deliver.
I think of agents as the next generation of software. They’re easier to create and can take action, which opens up new opportunities for automation. But that also makes governance and security essential. An agent without a clear owner, for example, can become a security concern.
We learned similar lessons with the Power Platform. Making it easier for people to build applications unlocked a lot of value, but it also created a need for ways to manage and secure what they were building. We’re applying those lessons to agents, helping you establish the right controls as adoption grows. The longer you wait, the more agents you may have to manage and the harder it can become to get everything under control.
➜ Governance helps organizations manage the agents they create. But the bigger opportunity is what those agents can do once they are connected to business systems and embedded in everyday work. That brings the conversation to another concept: headless business applications.
Q: What does “headless” mean, and how can it reduce manual work for employees?
James: I’ve talked with leaders across sales, service, and finance, and I’ve never heard one say their employees love keeping Customer Relationship Management (CRM) and Enterprise Resource Planning (ERP) systems up to date. People need that information in the system, but entering it manually takes time away from their actual work.
The idea behind headless applications is that you can interact with business systems without having to open the applications themselves. Instead of switching to a CRM to log a call or update a customer record, for example, an employee could use an AI tool to capture that activity and update the system as part of their existing workflow. With Work IQ, we’re working to connect AI to business data and actions to support these kinds of experiences.
Think about a field service technician finishing a job. Rather than spending time entering paperwork at the end of the day, they could dictate their notes on their phone and move on to the next appointment. The information still gets recorded, but the technician spends less time on administration.
To me, that’s the value of headless applications. You reduce context switching and manual data entry, allowing people to focus more of their time on the work they’re there to do.
➜ Reducing manual data entry is only part of the equation. For agents to coordinate work reliably across applications and business functions, they also need a consistent understanding of the underlying business processes and data.
Q: What role does Dataverse play in an AI-ready business?
James: I think one of the most important roles Dataverse can play is providing a common foundation for business processes that span multiple systems. You’re not just moving data into one place. You’re modeling how your business works so that people and agents can each handle different parts of a process, with a shared structure to help coordinate the work and maintain the right guardrails.
That becomes especially important as more people start building their own apps and agents. You can use AI to create a new system from scratch, but that doesn’t necessarily mean it’s the best approach. You still need consistent data, appropriate governance, and a way to manage what you’ve built over time. Without those foundations, even powerful AI tools can produce poor results when the underlying data isn’t reliable.
That’s why I see an important role for Dataverse and our low-code tools as businesses adopt more agent-driven automation. They provide a managed foundation for building solutions that can work across business processes, without every person having to create and manage the underlying infrastructure themselves. To me, that creates a more practical path to bringing AI into the business at scale.
A strong data foundation gives agents something reliable to work with. The next step is helping AI make better use of the information an organization already has, including the context spread across conversations, documents, and business applications.
Q: How can connected business data help AI deliver more useful results?
James: I think one of the biggest opportunities is helping AI connect information that’s already spread across your business. With Work IQ, we’re working to bring that context together so AI can do more than answer a question based on a single piece of information.
Eric, you mentioned how useful it can be to surface an email or meeting from years ago that provides important context for a current client conversation. When you combine that history with information in a relevant SharePoint site, you can get a much fuller picture of the customer and what’s happening with their business.
The more you use AI, the better it can begin to understand what you’re looking for and how you like to work. With Work IQ, we’re also looking to extend that understanding beyond individual users to business processes. To me, the opportunity goes beyond simply retrieving information. It’s about helping AI understand how work gets done and supporting more intelligent business processes. There’s a lot more to come in this area.
➜ Taken together, these ideas point to a broader shift in how businesses approach AI. The opportunity is not simply to create more agents, but to give people the tools, data, and governance needed to make those agents useful at scale.
Preparing for What Comes Next
Becoming agent-ready isn’t just about building more agents. It starts with empowering the people who understand your business to experiment, identify opportunities, and find better ways to work. As adoption grows, you also need the right governance, security, data, and processes to make those solutions reliable and manageable over time.
The technology is evolving quickly, and you don’t need to have everything figured out before getting started. Give people room to experiment, learn where AI can deliver value, and build on what works. At the same time, think beyond what you can build today to how you’ll manage and scale it over time. By combining experimentation with a strong foundation for governance and data, you can move beyond AI experimentation and turn its potential into lasting business value.
Before we wrap up, I'd like to thank James for sharing his insights and experiences. His perspective highlights the importance of combining experimentation with strong foundations for governance, data, and business processes as organizations become more agent-ready.
For More Information:
Are you ready to explore how AI agents can help your business improve productivity and streamline operations? Stoneridge Software can help you identify opportunities, connect AI to your business systems, and build a practical strategy for putting AI to work. Reach out to us to start a conversation about your business goals and next steps.
Under the terms of this license, you are authorized to share and redistribute the content across various mediums, subject to adherence to the specified conditions: you must provide proper attribution to Stoneridge as the original creator in a manner that does not imply their endorsement of your use, the material is to be utilized solely for non-commercial purposes, and alterations, modifications, or derivative works based on the original material are strictly prohibited.
Responsibility rests with the licensee to ensure that their use of the material does not violate any other rights.