Day two at Community Summit North America 2025 put Microsoft Copilot front and center, moving beyond the buzz to show how AI is reshaping work in real-time. From helping sales teams simplify processes and engage customers more effectively, to establishing AI-powered service flows that make customer support seamless, to driving productivity across Business Central and Fabric, the sessions revealed both practical wins and strategic imperatives. Copilot is becoming more than an add-on. It’s now a business necessity.
Fabric for AI Transformation in Finance
In an interactive session led by Craig Niemoeller, Director at RSM, and Lucas Orjales, AI Engineer at RSM US, “Reshaping Financial Services with Microsoft Fabric and AI,” the audience was guided through the journey a financial professional might undertake to complete specific tasks using the AI capabilities of Microsoft Fabric. The session began with a high-level overview of Copilot in Power BI, then shifted focus to automated model development with AutoML in Fabric. Finally, the presenters introduced Foundry, demonstrating how this extensive model library could be tapped to find a specific large language model (LLM) for customer engagement. Throughout each segment, Orjales provided a detailed guided walkthrough/demo.
Key Takeaways
- Specialist Use Cases: The session showcased how finely tuned Copilot in Fabric is for specific use cases, in this case, finance. It utilized language specific to the industry and effectively identified where to find relevant information. This specificity was highlighted through the creation of visuals, in the example used, churn probability, specifically using Copilot in Power BI.
- Automated Model Development: The speakers emphasized how easy it is to use Notebooks in Fabric to create machine-learning-driven predictive models. Users can automate every process, from data collection to prediction. There are out-of-the-box models for specific tasks, such as forecasting or binary classification. The churn model created was also incredibly easy to understand, thanks to clear visualizations.
- Integrated Toolkit: Furthermore, the pair demonstrated Fabric’s integrated AI toolkit, particularly how users can utilize Copilot in the initial stages of data discovery. This feature allows them to determine the exact data needed and classify it for building training datasets, which can then be used for model creation. As a final step, the model output can be taken to AI Foundry, where users can start taking action on the results surfaced, finding an LLM in Foundry specifically designed to address the issue and leveraging it to create a Power App. Ultimately, you don’t have to stop at predictions — you can take the next step, and all using Microsoft AI tools.
Turbocharging Sales Productivity
In another morning session “Copilot for Sales 101: Unlocking the Power of AI for Smarter Selling,” Scott LeFante of CongruentX showed how Microsoft Copilot is transforming everyday sales work. With AI summarizing emails, surfacing meeting insights, and centralizing deal data in Teams, sellers no longer waste time “hunting and pecking in their email.”
“Your reps can now focus on the conversation,” he said, while Copilot handles the rest.
LeFante also discussed the importance of training, security, and real-world adoption strategies. He reminded attendees that while Copilot delivers powerful automation and insights, its real value comes when reps know how to use it, trust the data, and apply it in the flow of work.
Key Takeaways
- AI That Works Where Reps Work: Copilot integrates directly into the tools sellers already use (Outlook, Teams, and Dynamics 365), making adoption frictionless. It automatically summarizes email threads, preps sellers with relevant meeting context, and syncs CRM data in real time, eliminating the need to toggle between platforms or dig for scattered information.
- Shift from Admin to Impact: With Copilot handling time-consuming admin tasks like note-taking, follow-ups, meeting summaries, and pipeline updates, reps can reclaim their focus. Instead of clicking through CRM fields or typing up recaps, they spend more time in conversations that actually move deals forward.
- Adoption Is Everything: The tech alone isn’t enough. Real success with Copilot comes from intentional rollout strategies. Teams need training on prompting, oversight for data accuracy, and metrics to measure impact. When sellers trust what the AI surfaces and see it saving time, adoption sticks, and value compounds.
GenAI: Lessons from the Field
In a series of engaging anecdotes, Dr. Ilyas Iyoob, a faculty member in Data Science and Artificial Intelligence at The University of Texas at Austin, made a compelling case for the transformative power of GenAI in various processes. This intimate discussion primarily focused on healthcare applications, showcasing the significant capabilities that GenAI has delivered as users transition from testing and implementation to fine-tuning and experimentation. Dr. Iyoob shared the experiences of practitioners who tackled persistent challenges and, alongside his team, developed impactful solutions by harnessing the power of GenAI.
Key Takeaways
- Don’t Get Left Behind: When we are solving real problems with real issues at scale, we need to be aware of the tools we have at our disposal because things are moving so fast. When all you have in your hand is a hammer, everything looks like a nail said Dr. Iyoob. His advice? Embrace the new and tailor it to your specific needs and requirements.
- Avoid Over-Contextualization: It’s also important to avoid over-contextualization. While AI models benefit from context, there are times when you don’t want the model to remember the previous context you provided it. Reset the context when you want separate results. Ensure you don’t maintain cache when you want to prevent the system from remembering past information.
- Stay Focused: With the speed of innovation, we are also witnessing a significant increase in complacency. Many people seem to think, “I’ll just put this here, and it will be fine.” However, if you remain disciplined, you can outpace the competition. By staying focused and consistently achieving more, you will be ahead while your competitors are languishing in their complacency.
AI Agent & Copilot Summit is an AI-first event to define opportunities, impact, and outcomes with Microsoft Copilot and agents. Building on its 2025 success, the 2026 event takes place March 17-19 in San Diego. Get more details.
Using AI to Cut Handle Times and Boost Customer Satisfaction
In ” Revolutionizing Customer Engagement with AI: Integrating D365 Contact Center and Custom Copilots,” led by Nancie Calder, Vice President and Global CRM Practice Lead at Avanade, and Will Hawkins, Lead Data Scientist and Founder at Ritewai, attendees explored how AI is transforming customer service through Copilots, GenAI, and intent recognition. The speakers unpacked real-world case studies, like Microsoft consolidating 16 support systems and 1-800-FLOWERS onboarding 6,000 agents across 10 brands in just six months.
One particularly engaging moment came during the Q&A, when an attendee raised a question about HIPAA and data privacy. Hawkins addressed it head-on, explaining Microsoft’s enterprise data protection and the ability to restrict Copilots to internal, tenant-secured knowledge only.
With demos of Copilot Studio, Power Pages, and Dynamics 365, the session delivered a hands-on roadmap for building intelligent, compliant, and agent-empowering support flows.
Key Takeaways
- Seamless AI-to-Human Handoff: As Hawkins demonstrated, modern Copilots now route customers based on true intent, not rigid decision trees. Escalations hand off cleanly to live agents, complete with sentiment data and conversation summaries.
- Knowledge Is Your Foundation: Nancie Calder stressed that even the most advanced AI can’t fix a broken knowledge base. Teams that structure content clearly and keep it updated see dramatically better performance from their Copilots.
- AI Should Empower, Not Replace: Both speakers emphasized that Copilots are digital colleagues — not replacements. They help agents stay focused on empathy and resolution by handling tasks such as note-taking, routing, and article suggestions in real-time.
Generative AI, Copilot in Business Central
In a fast-paced breakout session titled “Boost Productivity with AI: Unlocking the Power of Copilot in Business Central,” Michelle Serna, SVP of Revenue at @TruNorth Dynamics, walked attendees through the real-world impact of Copilot and GenAI within Microsoft Dynamics 365 Business Central.
Drawing from years of experience as an early Project AERA tester, Serna showcased how users — from finance teams to inventory managers — are using Copilot to streamline daily workflows: from reconciling multi-line bank statements and analyzing sales trends with pivot-ready visuals, to auto-generating item descriptions and sales orders based on email or CSV input.
The session moved fluidly between live demos and practical guidance on prompt strategy, responsible AI, and data quality. One standout moment came during a discussion on Copilot agents, where Serna fielded a question on handling multiple shared mailboxes in distributed sales teams. She explained how Microsoft’s Copilot agent preview now supports shared inbox automation — while emphasizing the need for governance, transparency, and proper licensing.
Key Takeaways
- Copilot Is Already Transforming Daily ERP Tasks in Business Central: Serna demonstrated how Microsoft Copilot is being used today to handle practical workflows like bank reconciliations, inventory analysis, sales order generation, and marketing text creation. The AI assistant reduces manual effort and accelerates decision-making, but still requires human oversight and clean data to perform effectively.
- Prompt Design and Data Quality Determine Copilot’s Success: The effectiveness of Copilot responses hinges on clear, layered prompts and high-quality input data. Serna emphasized that vague or overloaded prompts reduce accuracy, and poor internal data will lead to poor AI output — underscoring the need for prompt literacy and data hygiene.
- Responsible AI and Licensing Awareness Are Crucial for Adoption: The session stressed Microsoft’s principles of Responsible AI (fairness, transparency, reliability, privacy, and accountability) as essential guardrails.
Final Thoughts
Day two at Community Summit North America 2025 was pure momentum. Microsoft Copilot wasn’t just a buzzword — it was everywhere, reshaping how work gets done across sales, service, finance, and operations. And AI is no longer an experiment; it’s infrastructure. Financial professionals built predictive churn models in Fabric. Sellers learned to focus on conversations instead of admin work. Support teams watched Copilots cut handle times while protecting privacy and compliance. Every demo and discussion pointed to the same future, one where Microsoft AI tools help people do their best work faster, safer, and smarter.
Kieron Allen and Maya Rock contributed to this report.
Ask Cloud Wars AI Agent about this analysis
