Tools
Independent tools built to solve real pain points in AI/ML infrastructure and workflows. Each tool has its own site — this is the hub.
Builder Philosophy
- Every tool solves a real production pain point — not a synthetic demo problem
- Built by someone who has actually operated these systems at scale
- Minimal viable product first — ship early, iterate based on real feedback
- Each tool gets its own domain and site for focused UX
- Single-person team by design — lean, fast, opinionated
OptiFlow
AI Workflow Optimization
Optimize and streamline your AI/ML workflows. Intelligently routes tasks, manages resources, and reduces operational overhead for teams running complex ML pipelines.
LLM Cost Calculator
Inference Cost Modeling & Planning
Model the real cost of running LLM inference at scale. Compare cloud vs self-hosted, benchmark different model sizes, and plan capacity before committing spend. Built from real production cost data.
KV Cache Estimator
Memory Planning for LLM Serving
Estimate KV cache memory requirements for any model and serving configuration. Know exactly how many concurrent requests your GPU setup can handle before you deploy.
LLM Eval Harness
Automated LLM Quality Testing
Run systematic evaluation suites on your LLM applications. Catch regressions, measure hallucination rates, and build quality gates into your CI/CD pipeline.
Building in Public
I share the building process — what I'm working on, what's working, and what's not — on Substack and X. Subscribe or follow to watch these tools get built.