Learn it by building it.
Live, hands-on workshops that take a real problem end to end. You leave with the working artifact, not just slides. Each workshop runs on a private environment; your Spicy Data contact gives you the access credentials.
Claude Foundations: Solo to Team
Get a team from first prompts to a shared, quality-gated skills marketplace. Day 1 builds solo reliability in the Claude desktop app: the right surface for each task, lean context, and repeated work turned into skills and scripts. Day 2 merges everyone into one GitHub skills marketplace with size lint, LLM review, and human review on every contribution.
Context Engineering for Omni
For teams new to Omni or evaluating it. Across two days on a provided demo dataset, you turn a raw semantic model into one an AI answers correctly: document fields, write sample queries, build an eval set, tune the cost-quality dial, and watch the Omni Agent flip from wrong to right at every step. No Omni experience assumed, nothing from your own company required.
Context, not the model, is the lever: on Anthropic's own internal analytics agent, this work moved accuracy from about 21% to 95%+ without touching the model. Omni's semantic layer is the native home for that context.
You leave with:
- A curated Omni Topic you built and tested yourself, with a measured before-and-after pass rate
- The full context-engineering method, from scope and grain through evals, trust, and governance
- Your company worksheet: your first Topic, trap fields, ten questions, and adoption sketch, drafted by you
- Cheat-sheet cards and an Omni docs map for everything you touched
Want a shorter look first? The mini workshops take one concept and prove it live in 60 minutes, free.
See mini workshopsMore workshops are on the way. Want one run for your team? Get in touch.