Skip to content
View maxgregaabig's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report maxgregaabig

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
maxgregaabig/README.md

Max Regan — dependable AI, from model to production

Hi, I'm Max.

Software Engineer · AI Systems & ML Infrastructure

Selected projects

Agent execution, retrieval & inference

durableflow — What survives when an agent crashes?
A Python/SQLite lab for checkpoints, persistent approvals, model fallback, and failure injection.

pgvector-rag — Retrieval with inspectable evidence.
A FastAPI document-search service with PostgreSQL/pgvector, Redis caching, chunk-level citations, and an evaluation harness. Deterministic providers make the core behavior testable without API keys.

model-serving-stack — Serving is a systems problem.
A reference environment for vLLM, Triton, and Ray Serve, with Kubernetes deployment configurations and GPU observability. The FastAPI gateway targets vLLM; alternative runtimes have separate deployment paths.

autoresearch — A bounded loop for autonomous experimentation.
An agent edits training code, runs a fixed-budget experiment, evaluates the result, and keeps or reverts the change. The interesting boundary is the experiment contract: what may change, what stays fixed, and what counts as improvement.

Developer tools

Project What it makes visible
FlakeLedger Flaky tests across reruns of the same commit, with a transparent model for retry compute and developer waiting costs. Python; offline JUnit XML analysis.
ReviewTide Review waiting times and code-ownership concentration from PR and Git exports. Reports sample sizes and process bottlenecks without ranking individual productivity. Python; offline analysis.

Pinned Loading

  1. FlakeLedger FlakeLedger Public

    Flake ledger for JUnit runs - classify flaky versus broken, quarantine with evidence, and price the retry cost across CI runs.

    Python 20 3

  2. ReviewTide ReviewTide Public

    Review latency and ownership analyzer for git logs and PR exports - where review time goes, who the bottleneck is, and how the queue moves.

    Python 7 1

  3. autoresearch autoresearch Public

    Python 1

  4. durableflow durableflow Public

    Python 1

  5. generative-ai-node generative-ai-node Public

    JavaScript 1

  6. pgvector-rag pgvector-rag Public

    Python 1