1. Does the 400 cover access to the platform and if so, for what duration and what usage quotas, etc. 2. Can we purchase access to the platform without the course, e.g let us play with it for N months (course seems aimed at junior level devs)? 3. The tools and method, as described, don't mention any tools for minimizing divergence/conflation/hallucination and maximizing repeatability/verifiability etc. Do you have a prompt optimization tool (ala DSPy), a built-in pipeline that uses one LLM to cross check another, or that uses deterministic tools to verify output of LLM (see Campbell, Wimsatt heuristic for managing/triangulating in uncertainty)? I'd be more interested in this new methodology if you were tackling head on the inherent limitations of LLMs (e.g. Yann LeCunn's vision)