Watch

Video and demos.

  1. Ongoing

    Fun Demo Fridays

    YouTube

    Small, hands-on demos exploring what is possible with AI and developer tools.

  2. May 2025

    Learn Live: Integrate custom tools into your agent

    Microsoft Reactor

    A live walkthrough of giving an agent custom tools with Microsoft Foundry and Azure AI Agent Service, so it can call out to real services instead of guessing.

  3. Apr 2025

    Building an AI agent with Prompty and Microsoft Foundry

    Microsoft Reactor

    Building an agent that triages and tags GitHub issues, using Prompty for structured prompt engineering and Microsoft Foundry to run it.

  4. 2025

    Azure AI Fundamentals, 2025 edition

    Microsoft Reactor, AI-900 series

    A full series covering the Azure AI fundamentals, from what the services actually do through to sitting the AI-900 exam.

Workshops and labs

Hands-on sessions I have written or delivered, most built to be run at your own pace.

  1. Jul 2026

    Biashara Agent Workshop

    Safaricom Decode

    Build a bilingual English and Swahili AI agent for a Nairobi restaurant using the Microsoft Foundry Toolkit in VS Code. Covers agent setup, system prompts, MCP tools for grounded answers, evaluation, a Flask chat interface, and exporting to production code with Microsoft Agent Framework.

  2. Jun 2026

    Foundry Toolkit and Hosted Agents Workshop

    Microsoft Foundry

    Build, test, and deploy single-agent and multi-agent Hosted Agents to Microsoft Foundry from VS Code, using the Foundry Toolkit and Microsoft Agent Framework in Python.

  3. May 2026

    LAB521: Improving Agent Behavior Using Reinforcement Learning from Traces

    Microsoft Build 2026

    Agents often work, but prompts alone make their behaviour hard to control. This lab starts with a working agent, inspects its execution traces, defines good behaviour using evaluation graders, and applies reinforcement fine-tuning to reinforce better decisions without redesigning the agent.

  4. AI Tour 2025/26

    BRK441: Build and Launch AI Agents Fast with Foundry Toolkit for VS Code

    Microsoft AI Tour

    Prototype with open models in the Foundry Toolkit Playground, iterate on prompts in Agent Builder, evaluate against a test dataset, then take the same work from prototype to a deployed, monitored agent.

  5. AI Tour 2025/26

    BRK443: Efficient Model Customization with Microsoft Foundry

    Microsoft AI Tour

    When does fine-tuning actually make sense? This session walks through the customization options in Microsoft Foundry, then gets hands-on with synthetic data generation, distillation for cutting cost, and SFT for shaping tone.

  6. Nov 2025

    PREL13: Observe, Manage and Scale Agentic AI Apps with Microsoft Foundry

    Microsoft Ignite 2025

    An end-to-end lab on taking an agentic app past the demo stage: instrumenting it, watching what it actually does in production, and scaling it with Microsoft Foundry.

  7. 2023

    Analyzing Snapshot Serengeti data with Microsoft Fabric

    Microsoft Open Awesome Workshops

    A self-paced workshop using camera trap data from the Serengeti: load it into a Lakehouse, query the SQL endpoint, visualize in Power BI, then transform and analyze the images with Apache Spark in Fabric.

Curricula and courses

Longer-form programs and learning paths.

  1. Ongoing

    Generative AI for Beginners

    Microsoft

    A course of lessons teaching everything you need to start building generative AI applications.

    Read the RAG and vector databases lesson

  2. Ongoing

    Using GitHub Copilot with Python

    Microsoft Learn

    A module on pairing with Copilot in Python: writing prompts that work, reviewing what it suggests, and knowing when to push back.

  3. Ongoing

    AI for Beginners

    Microsoft

    A 12-week curriculum covering the foundations of AI, from symbolic approaches through neural networks, computer vision, and natural language processing.

  4. Ongoing

    Prompty documentation

    prompty.ai

    Documentation for Prompty, an asset class and format for structuring LLM prompts so they are observable, understandable, and portable across tools.

  5. Aug 2026

    Alliance Girls High School Software Engineering Program

    with TOFA and Alliance Girls Alumni

    A project-based software engineering program for Grade 10 students. Term 2 covers foundations, careers and resources; Term 3 is a group project sprint ending in a showcase. Includes a facilitator guide, icebreakers, demos, and beginner career-path guides.

  6. 2022

    30 Days of Data Science

    Microsoft 30DaysOf

    A month-long roadmap of data science learning resources, samples, and curricula, structured so you can follow it a day at a time.

Talks

Sessions and community talks.

  1. Jan 2022

    Manipulating and Cleaning Data

    Microsoft Reactor Tel Aviv

    Real-world data is messy. A beginner-friendly session on missing and duplicate data, joining and merging datasets, and exploring your data visually and statistically before you model anything.

    Watch the recording

Read

Articles, tutorials, and community reflections.

  1. Ongoing

    Microsoft Community Hub

    All posts

    Articles, tutorials, and community reflections published through Microsoft Community Hub.

  2. May 2025

    Is it a bug or a feature? Using Prompty to automatically track and tag issues

    Microsoft Developer Community

    Triaging issues is the kind of chore that quietly eats a maintainer's week. This walks through building a Prompty agent that reads an issue and tags it for you.

  3. Apr 2025

    Automating PowerPoint generation with AI

    A Learn Live series case study

    What it took to automate slide generation for a Learn Live series, and where the automation earned its keep.

  4. 2025

    Translating AI and ML for beginners curriculums in less than a day

    Co-op Translator

    Translating two full beginner curriculums into multiple languages in under a day, using Co-op Translator.

  5. Mar 2023

    Using Azure ML to train a Serengeti data model for animal identification

    Open Data Science

    Training a model on Snapshot Serengeti camera trap images to identify the animals in them, from data prep through to a deployed endpoint on Azure ML.