Workshop: Analyse your Receipts with Tensorix

Build a private AI app in one evening

Sometimes you do not want to send your data away to someone else's cloud. You want private AI on infrastructure you
can rely. Tensorix (www.tensorix.ai) offers: 30+ open-weight models (DeepSeek, Qwen, Kimi and
friends) on EU-sovereign infrastructure with no data retention, behind a drop-in OpenAI-compatible API.

This is a hands-on workshop. Bring a laptop. You will leave with:

• A working app you built yourself, running on your laptop
• A pipeline that extracts text from images
• A receipt classifier you can show your colleagues, and the skills to extend it
• A Tensorix account, set up and loaded with free credits (pay-as-you-go after that, no monthly fees or minimums)
• Pizza eaten

On the night I will step through building a Streamlit app using spec-driven development, connecting it to Tensorix,
and getting your colleagues using it. No AI experience needed; a little Python helps. I code in cursor so downloading that will help. But if you know another ide you will be fine to use that.

If you can sign up at https://tensorix.ai/ for an account that will save some time.

Tensorix are hosting us at their space and are providing the LLM credits and the pizza. It is a workshop with limited numbers so please keep your rsvp accurate if you can.