Working labs

Read the method, inspect the evidence, then run it yourself. Notebook previews show saved outputs; they do not run an LLM in your browser.

Read Lab 01 in the notebook viewer →
  • Lab 01: Continuous risk identification github.com

    Build an evidence pipeline that connects mock system records to business context and candidate risk statements. Identification only, with human review.

    Risk identificationPythonJupyterLLM