AI Knowledge Platform
In ProgressModular, cloud-native AI knowledge platform on AWS enabling semantic search and grounded question-answering across structured and unstructured documents — reusable foundation for multiple AI applications (RAG demo, technology radar, regulatory monitoring).
Technologies
Problem
Organisations need a cost-efficient AI system that processes diverse document formats, retrieves contextual answers with transparent source attribution, and scales from a public demo to private knowledge management — without vendor lock-in.
Approach
Layered architecture with separate ingestion, processing, storage, retrieval, and LLM layers. PostgreSQL + pgvector (HNSW index) for vector storage, SHA-256 deduplication to avoid unnecessary re-embeddings, abstracted LLM providers (OpenAI/Anthropic) for vendor independence. Cost controls built in as a first-class requirement.
Result
Reusable platform library with FastAPI endpoints, CLI, 18 unit tests (day 1), strict type checking (mypy) and a complete local development environment via Docker Compose. Phase 1 scaffolding complete.
Learnings
Python package names must not shadow stdlib modules (platform → aiplatform); Alembic requires synchronous drivers (psycopg2) while FastAPI uses async (asyncpg) — two separate connection strings solve this; HNSW index in pgvector requires no training phase; cost controls must be built into core infrastructure early, not added later.
Relevance
Demonstrates complete data pipeline thinking (ingestion → preprocessing → vector storage → retrieval → LLM response generation), practical AWS cost management, and the ability to design and implement a reusable, layered AI platform from scratch.