An open, automatically curated Map of Content (MOC) grouping working notes, essays, and artifacts across all clusters.
01. Buildss
Hardware experiments, firmware, and software artifacts.
- RunSarcasm GPS Tracker — A custom hardware/software project building an aggressively sarcastic GPS tracker for marathon…
- Migrating this site to a block theme, from scratch — Rebuilt koushik.site as a minimal custom FSE block theme instead of a…
- Training data dashboard, mostly to stop checking four apps — Pulled GPS, pace, and heart rate data into one local dashboard instead…
- A weekend RAG eval harness, built out of spite — Got tired of trusting vibes on whether a retrieval change helped or…
- RunSarcasm: added track-mode distance correction — Spent the weekend porting the track-mode correction logic into RunSarcasm's firmware. Turns…
02. Librarys
Curated reads, book citations, and external references.
- Stripe’s Design Culture — An excellent deep dive into how Stripe treats documentation as a first-class…
- The Complete Guide to Enterprise AI Governance in 2026 — Dense but useful. The six-step framework is standard stuff, the value is…
- the5krunner on GPS Accuracy in Sports Watches — The most rigorous independent testing of running watch GPS I've found. The…
- Top 5 LLM Evaluation Frameworks Compared — Good breakdown of when to reach for Ragas versus a full observability…
- Meadows: Leverage Points, Places to Intervene in a System — The original essay, not a summary of it. Read this once a…
03. Logs
Field observations, quick takes, and thinking out loud.
- Read: The Anatomy of Technical Due Diligence — Great point on unit economics. A novel solution that improves performance by…
- Read: Transforming Systems with Leverage Points — Fair critique that Meadows' hierarchy lacks empirical grounding, she said as much…
- Read: Inside Stripe’s Engineering Culture — The operational excellence section is the part worth re-reading, not the docs…
- Read: Postgres Won, the Complete 2026 Guide to Vector Databases — The 471 QPS number is real and I've reproduced something close to…
- Read: How Product is Changing in 2026 — Agree with the framing here, but I'd push it further: evaluation isn't…
04. Notess
Synthesized essays and atomic working notes.
- What Breaks When LLMs Enter Regulated Enterprise Workflows — Most enterprise AI initiatives do not fail on raw model intelligence. They…
- The cost of abstractions — Just spent the afternoon untangling a "helpful" ORM layer that was executing…
- Shadow AI is just a feedback loop nobody built — Talked to a few people this week who admitted to pasting confidential…
- Track mode is the GPS feature nobody talks about — Ran intervals on the track this morning wearing two watches, one with…
- pgvector at 50M vectors, and why I stopped recommending Pinecone by default — The pgvectorscale benchmark everyone is citing this year: 471 QPS at 99%…
- A tub, a bank account, and every system I manage — Re-read the Meadows bathtub explanation for the third time this year. Stock…
- Leverage Points, and Where I Keep Pushing the Wrong One — Donella Meadows' list of places to intervene in a system has one…
- Product Judgment Is the New Moat — The AI coding and drafting tools got good enough this year that…