Data & Analytics
SPEC.md · 7 documents
scope
Defines a Python tool that downloads YouTube transcripts and comments, then runs a 3-stage AI pipeline to produce topic summaries and atomic insights with hybrid search.
deep-cuts — Media Intelligence Pipeline
Defines a pipeline that ingests YouTube shows and podcasts, transcribes them, runs LLM analysis, stores results with semantic search, and generates infographic cards per episode.
Infinite-Context Chat Storage Specification
Specifies a graph-database-backed chat storage system that uses vector embeddings and adaptive indexing to give LLMs unlimited context without token waste.
Vector Database Shootout - Functional & Technical Specification
Defines a benchmarking suite to compare performance of five vector databases across embedding models, dimensions, and query patterns.
Software Carbon Intensity for AI Specification
Extends the SCI methodology to AI systems, defining how to measure and report carbon emissions across AI lifecycles and personas.
Logging & Evals
Documents a trigger-based logging system, cost tracking, token analytics, and a separate evals dashboard for reviewing LLM interactions.
GPU Selection Guide for Large Language Models (LLMs)
Guides GPU selection for LLM inference, fine-tuning, and training by mapping model sizes, precision levels, and budgets to VRAM requirements.