This is a hands-on, build-heavy day for technology and business leaders, IT professionals, and knowledge workers across local and state government. Rather than a morning of theory followed by an afternoon lab, the day opens with just enough grounding to make sense of what you’re building, then spends the bulk of its time in Ejento AI itself — building a working agent, live, that gets more capable exercise by exercise.
What you’ll walk away with
Participants leave with a clear picture of what a modern agentic platform is actually made of — enough to cut through vendor claims and ask the right questions — along with a working AI agent, built by hand in their own Ejento AI sandbox, that grows more capable through the day.
The day also lays out a practical playbook for scaling adoption safely, covering what to automate, what stays human-in-the-loop, and when to loop in Legal and InfoSec, and leaves the room with the vocabulary and judgment to hold their own in any AI conversation with vendors, IT, or executives.
Curriculum and topics, by session block
1. Setting the stage — 90 minutes
The day opens with one master infographic, and every core concept gets a tight, one-to-two-slide treatment anchored to that same diagram: large language models (LLMs) and what they are and aren’t, prompts and skills, context (RAG, MCP, and tools), guardrails, evaluation, access control (ReBAC), the agent harness, observability, and security, governance, and compliance. The room builds one consistent mental model instead of a stack of disconnected terms.
That same diagram is then redrawn as Ejento AI’s own architecture, making the point explicit that Ejento is built entirely on Microsoft Foundry and Azure, and is placed directly alongside Copilot Studio, Microsoft 365 Copilot, and Agent 365 so participants leave with a clear-eyed view of where each tool fits.
2. Building, live — the rest of the day
Every participant sets up their own sandbox and builds a chat agent against a shared HR-benefits scenario, so the room stays in sync, then progressively adds new capabilities to that same agent, exercise by exercise, so it grows in front of them rather than being rebuilt from zero each round.
3. Closing the loop — wrap-up
The day closes with how to expand adoption from one project to a broader rollout, when and how to loop in Legal and IT, and what’s coming next in AI-related compliance.
Format
One full day, 9:00 AM – 4:00 PM Central, at the Microsoft Innovation Hub in Houston — officially the Microsoft Technology Center — 750 Town and Country Blvd, Suite 1000, Houston, TX 77024. Roughly 90 minutes setting the stage, and the rest of the day building live.
Instructor
Raja Iqbal brings 20 years in AI and a PhD in pattern recognition. He is the founder of Data Science Dojo and Ejento AI, adjunct faculty at the University of Pittsburgh, and was formerly at Microsoft.