Here lie ideas that were... ahead of their time. Or behind. Mostly just behind.
- prof_ai - For when you want your profiling to take 100x longer and be 0% more helpful.
- PoolX - I got really, really excited about not allocating memory. It was a phase.
- Memora - A management system for multiple PoolX instances. It solved the problem of "what if you need a second PoolX".
- TokenSpan - A "revolutionary" prompt hack. It was neither revolutionary, nor much of a hack.
- GoCortex - Taught an AI to be my intern. It spent most of its time generating excuses.
- IdleSpy & Yieldpoint - Incredibly niche solutions for problems that may not actually exist.
- PromptMash - I built agent pipelining. Then I discovered n8n.
- GhostDev - A "first-mover" in the emergent autonomous agent orchestration space.
- I dissect my projects—both the hits and the misses—as technical experiments. I write about the insights on Medium and discuss them on Reddit.
- I like treating everything as an experiment, so you're just doing me a favor when you tell me that I am wrong.
Feel free to reach out:
