Build AI-first delivery on Cloud and AI Platform.
A 60 to 90 minute partner workshop for Azure delivery teams. Walk through the lifecycle of an AI-first project, see the toolbelt your consultants get, and leave with the material to run it inside your organization.
A working delivery motion, not a webinar.
Agents and MCP-connected tools are starting to handle the parts of delivery that used to eat consultant hours. Backlog grooming, scaffolding, infrastructure, test, the admin around every project. The partners building this motion now have a real window.
This workshop walks you through the lifecycle of an AI-first project on Azure: client brief, planning, demo, build, ship. You see the actual toolbelt the field is using on Cloud and AI Platform: skills, agents, MCP servers, and the GitHub Copilot CLI front end that ties it together.
Everything is on GitHub. Decks, scenario, exercise, setup, FAQ. Clone once and keep it after.
Everything to run it.
Pick a card. The deck is the source of truth. Exercise and setup live in the repo.
Exercise scenario
Zava, an industrial OEM. Client brief, target architecture, deliverables, and the full workshop scaffold.
Open exerciseSetup tiers
Three setup options based on how deep you want to go in the room: GitHub Copilot CLI alone, Copilot CLI plus Azure CLI, or the full agent stack.
Setup overviewStarters
Pre-baked CLI configurations to drop into a partner consultant's machine. Skills, agents, MCP servers wired up.
Open startersFAQ
Common questions from past sessions: licensing, what to bring, what runs where, what's missing, what's next.
Open FAQThree depths. Pick the one your room needs.
The deck lands regardless of setup. Hands-on depth changes based on what your consultants have ready.
Run through the deck and the scenario as a discussion. No installs. Best when you want a leadership-level conversation.
Each consultant has GitHub Copilot CLI. Walk the exercise live, see real CLI sessions, no Azure deployment.
Copilot CLI plus Azure CLI plus the agent and MCP servers from the starters. Build, deploy, ship inside the session.
Clone once. Keep it after.
Everything for this track lives in the public repo. Clone it, branch it, run it inside your organization.
Clone the public repo
The CAIP track ships under tracks/caip/. Or download the whole tree from the GitHub Code button.
git clone https://github.com/JW-Sthlm/frontier-consultancy-public.git
Questions, ideas, edits?
Track owner: Johan Wallquist (Microsoft, EMEA Partner Solution Architect, Data and AI). Reach out via Teams or email for anything CAIP-specific. For the program overall, see the all-tracks page.