Built for Signal. Not Noise.

After years in crypto markets — first as a retail trader, then as a more systematic one — I kept seeing the same pattern. Traders with sound market reads making consistently bad entries because their analysis was overridden by sentiment, social pressure, or FOMO at the worst possible moments.

Twitter sentiment would override volume analysis. A celebrity endorsement would negate a clean bearish divergence. A euphoric crowd would mask the funding rate signal screaming “caution.” The information was there. The problem was aggregation and discipline.

I started building a structured framework to solve this — one that forced systematic factor weighting before any directional bias was formed. That became InDecision.

The InDecision Framework aggregates six independent market signals — each measuring a distinct dimension of market structure — and synthesizes them into a single direction and conviction score. No single factor dominates. The model abstains when the signals conflict.

The name is intentional. “InDecision” isn't a bug — it's a feature. The framework is designed to force genuine indecision when the market genuinely doesn't have a clear setup. A forced call under uncertainty is worse than no call.

I'm a software engineer and systems thinker who got deep into crypto markets after realizing that most “analysis” in the space was post-hoc rationalization dressed up as insight. I wanted a model, not a vibe.

I built InDecision as a personal trading tool first. It went public because the signals were consistently good enough that sharing the framework — not just the calls — seemed more valuable than keeping it private. The goal is to teach the model, not sell picks.

Outside of InDecision, I run jeremyknox.ai — where I write about AI engineering, systems thinking, and building in public.