Free resources

    Free workshops, courses, and interactive tools on shipping production AI and ML, built for working data scientists. No fluff, just the parts that get you from prototype to production.

    Workshops and courses

    Recorded workshops, a deep dive, and a five-part course by email.

    Three recorded workshops as a numbered timeline: an AI analyst in Claude Code, connected to your data with MCP, then to your business with a semantic layer.
    Free workshop seriesWatch on demand

    From Chat Window to AI Analyst

    Three recorded workshops: an AI analyst built in Claude Code, connected to your database with MCP, then to your business with a semantic layer. Demo files included.

    Watch the series
    The five stages of the MLOps course as a pipeline: structure, serve, track, automate, monitor.
    Free 5-part course

    MLOps for Data Scientists

    Take your models from a notebook to a monitored system running in production. Five parts, each a practical guide paired with a working code repo, delivered to your inbox one part a day.

    Get the course
    Architecture diagram: a question flows through routing logic and a semantic model to the warehouse, wrapped in guardrails, with an observability layer underneath.
    Free workshop

    Agentic Analytics for Data Scientists

    A 30-minute deep dive on shipping production agentic analytics: the architecture, the trade-offs, and the parts the tutorials tend to skip.

    Get the workshop

    Interactive tools

    A production ML system you can click through: the system, the repository, and the months after launch.

    All tools
    The four stages of a production ML system as connected boxes: make the model, ship it, run it, watch it, with a feedback loop back to the start.
    Free interactive

    Anatomy of a Production ML System

    A clickable map of a real production ML system, end to end. Open any component for what it does and why it exists, play a flow to watch data move through it, or step the build order to watch the whole thing assemble.

    Open the map
    A production ML repository file tree with folders color-coded by the job they do, and data and model folders marked gitignored.
    Free interactive

    Anatomy of a Production ML Project

    The same system seen as a repository: the folder layout MLOps-mature teams converge on, with every file and folder clickable for what it holds, why it exists, and where it shows up in a running system.

    Open the layout
    A reference distribution beside a current one that has shifted, over a strip of daily drift checks with a run of them flagged, and two conversion-rate lines.
    Free interactive

    Data Drift Simulator

    Six months of a deployed model's life, played day by day. Nine short steps walk you through what a drift check sees, when it fires, when it fires for no reason, and what it can never see. Then every setting is yours to try.

    Open the simulator
    A Slack question flowing through four phases, ask, understand, query and reply, into a warehouse and back as an answer, with definitions and traces beside the path.
    Free interactive

    Agentic Analytics: System Map & Roadmap

    Follow a business question from Slack to SQL and back through an agentic analytics system. Open any component, watch a convincing answer go quietly wrong, then follow the learning roadmap from text-to-SQL to a trustworthy analytics agent.

    Open the map
    A results table of eval cases with ID, question, expected and observed values and a pass or fail mark, beside a large score of 53 out of 60 cases, with an evidence strip of trace steps under the table.
    Free interactive

    Evals, Step by Step

    Plan the cases, graders and run budget for your own AI agent's evaluation, and learn each decision on a simulated Slack analytics bot, all in your browser.

    Open the workbench