Production systems and LLM agent infrastructure: distributed pipelines (Dask, Celery), agent observability (Arize Phoenix), Python, Django, task queues, security gates, and Linux workflow. Experiment artifacts and engineering notes below.
Browser-only event planner: you say how to rearrange your day in plain language and it rewrites the calendar. The model never touches state — it returns a small set of operations, and deterministic code resolves what you meant, validates the result and shows a before/after diff you approve. No backend, bring your own key, and it ships its own eval harness across Claude, GPT and Gemini with per-case pass/fail, latency, tokens and cost.
CLI framework for declaratively defining agent flows, running evals, and caching tokens between test runs. Combines user-flow testing and token cost optimization into one tool — same execution trace, cache aggressively, measure quality simultaneously.
Custom typestubs that track payload annotations by source to gate unsafe ML deserialization (pickle, HDF5) at the type-checker level. Security research from huntr.com CVE work on serialization-route vulnerabilities in GenAI platforms. Lean 4 soundness proofs.
~300 LOC task queue on Django ORM — prototype to production. Covers memory leaks from fork, pessimistic locking for exactly-once delivery, and security hardening. No Redis, no Celery: just the database you already have.
Architectural approach for making source, operation, risk, and policy state explicit through typed side channels rather than text-only prompts. Improves model alignment without prompt bloat by encoding context in the type system.
Systems for high-performance ML research: HF-streaming for large artifacts and the dual-emit data-driven paper pattern.
Pickle is a CVE factory. falcon-secure uses Python type stubs and Lean 4 soundness proofs to gate unsafe deserialization at the type-checker level.
Building a ~300 LOC task queue on Django ORM from prototype to production — memory leaks, fork pitfalls, pessimistic locking, and security hardening.
Kitty terminal, kittens, shell integration, Starship prompt, and turning the terminal into a complete development environment.
Agent evals that actually catch regressions: a Python flow/expect DSL for deterministic assertions, Arize Phoenix for fuzzy semantic evals, and cache-cluster grouping for token savings.
Custom status lines for ambient awareness and git worktrees for parallel AI-assisted development sessions.
Diagnosing and fixing SSH latency over high-latency VPN links, with focus on the TCP-over-TCP problem and connection multiplexing.