ADR-004: Dual Stream-Processing Engines — Faust (Python) + KSQL
· 3 min read
ADR-005: Kafka REST Proxy for the Weather Producer
· 2 min read
ADR-006: Tornado Async Web Server for the Real-Time Transit Dashboard
· 2 min read
Architecture Decision Records
· 2 min read
Welcome to the JigsawFlux Blog
· One min read
Welcome to the JigsawFlux Blog — a space for updates, ideas, and stories from the JigsawFlux open-source community.
Running a Local LLM with Ollama and MCP — An Architecture Spike
· 10 min read
AI inference doesn't have to mean a cloud API call. This post walks through a spike I built to run a locally hosted language model through a clean, layered architecture using Ollama and the Model Context Protocol (MCP).
The full source is on GitHub. Contributions and feedback welcome.