Free Live Workshop Series for Data Scientists

From Chat Window to AI Analyst

Three live builds. An AI analyst that runs real analyses, queries your database, and gets business questions right. See what it takes, then start your own.

Sep 24 to Oct 8, 2026Live online, 5:00 PM UTCRecordings for all registrants

Join 0 data scientists already registered

Free. One registration covers all three sessions.

What happens in each session

Something breaks on screen, the fix gets built live, and you get the files afterwards.

1
Sep
24
Claude Code for Data Scientists
Your first real analysis with an AI analyst
Thu, 5:00 PM UTC · with Markus Kuehnle
Most of us use AI in a chat window, and the real work still happens by hand afterwards. This session shows what changes when the AI works inside your project instead: it runs the analysis, produces the charts and the write-up, and you learn how to tell when its work can be trusted.
You leave with: a Claude Code setup that takes your own data from question to finished analysis, and a way to check its work before you trust the output.
2
Oct
1
Connect Your AI Assistant to Your Data
Query your database in plain English with MCP
Thu, 5:00 PM UTC · with Adel Hassen
An AI that can't reach your data can only guess. Adel walks through the ways people try to fix that, why most of them fall short, and how MCP gives your assistant a real connection to your database. Then you ask it real questions and see both how far that gets you, and where it stops.
You leave with: an understanding of what MCP is and what it takes to connect AI to your data.
3
Oct
8
An AI Analyst Your Business Can Trust
The semantic layer between your data and every AI tool that touches it
Thu, 5:00 PM UTC · with Markus Kuehnle
Inside a company, the hard part isn't the model. It's that an AI with full access to your data still doesn't understand your business, so it answers confidently and wrong. Markus shows how context engineering fixes that with a semantic layer between your data and every AI tool that uses it. You learn how to structure one, how to measure whether a talk-to-your-data agent is actually right instead of trusting it, and where the layer sits in a real company's architecture. The same agent goes from mostly wrong to nearly always right.
You leave with: a clear understanding of context engineering and the semantic layer, and a fix you could bring to your own warehouse in a day.

Who this is for

You use ChatGPT or Claude every day, and most of it is paste, prompt, copy, repeat.
You've pasted a schema into a chat and hoped. You haven't connected AI to a real database yet.
Job postings list MCP and agentic workflows, and you want to see them on a real dataset, not a slide.
You've finished courses that changed nothing about your Monday.

Your experience is the asset. Knowing what "engaged user" means in your business is exactly what the AI is missing.

Your hosts

Andres Vourakis

Sr. Data Scientist, Founder at Future Proof DS

8+ years in tech and applied AI/ML. Creator of the AI Workflows Bootcamp. Co-hosts all three sessions.

Markus Kuehnle

AI Engineer @ Wieland

AI Engineer in Supply Chain at Wieland AG, based in Ulm, Germany. Builds ML models and AI systems that run in production, and teaches AI systems on LinkedIn and in schools. Leads the live builds in sessions 1 and 3.

Adel Hassen

Data Scientist @ Dataiku

Experienced in applied AI and ML systems, with a focus on driving real business impact. Leads the live build in session 2.

See what sits above the chat window

Three live builds, from a chat tab to an AI analyst that knows your data and your business. Every recording and file lands in your inbox.