Projects
What we're building and shipping.
Traceprop: End-to-End Provenance-Guided Data Attribution for Auditable Machine Learning
We built Traceprop because every ML pipeline we audited had the same gap: source files on one side, predictions on the other, nothing connecting them. Traceprop closes it with three layers - ProvenanceTensor lineage (sub-1% overhead at 1M elements), JL-projected GradientStore attribution (LDS 0.622 on tabular, 266x faster than TRAK), and provenance-guided unlearning that exceeds retrain-from-scratch. EU AI Act Article 26 enforcement starts August 2026. Apache 2.0. Two lines of code change. Everything else stays the same.
SynapseKit v1.7.0 - Lightweight LLM Framework
A production-grade LLM framework built because LangChain frustrated us,and we decided to measure whether a simpler approach could actually work better. SynapseKit v1.7.0 has 2 dependencies (vs LangChain's 67), a 30x faster cold start (12ms vs 360ms), and built-in cost guardrails that prevent a single agent run from blowing your API budget. Chains, agents, RAG pipelines, and tool use,all with zero magic and full debuggability. When something breaks at 3am, you can read the source in 20 minutes. MIT-licensed, fully documented, and battle-tested across 18 objective benchmarks.
LLM Framework Showdown,30 Notebooks on Kaggle
A reproducible benchmark comparing SynapseKit, LangChain, and LlamaIndex across developer experience, RAG pipelines, agent capabilities, and production concerns. Every notebook includes methodology, raw results, statistical significance tests, and our interpretation. No hidden preprocessing, no curated datasets. If you disagree with a result, fork the notebook and prove us wrong,that's the point. We measure cold start, memory footprint, streaming latency at P99, and 15 more dimensions that matter in production.