Skip to content
View Tmwakalasya's full-sized avatar
  • dc | sf
  • 17:53 (UTC -04:00)

Block or report Tmwakalasya

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
Tmwakalasya/README.md

Tuntu Mwakalasya

Just love computers, man.

I build backend systems and developer tools, mostly in Go and Python. I’m interested in what happens when things fail: a dropped request, an interrupted write, a backup that won’t restore.

FIU Computer Science · Class of 2026

LinkedIn · Merged contributions


Patches upstream

Uber SubmitQueue — CI and merge orchestration

Traced failures from Git and GitHub through the orchestrator’s retry logic, then fixed places where error information was getting lost.

  • #774 — wired in HTTP error classification so transient responses such as 429 and 503 reach the retry policy.
  • #781 — preserved Git error chains and cancellation identity, and registered the Git classifier.
  • #806 — preserved GitHub API status errors so temporary metadata-fetch failures can be retried.

OpenAI CLI — command-line tooling

  • #231 — fixed Zsh completion when the cursor is before trailing arguments; added regression coverage using the rendered shell script.
  • #186 — eliminated shared mutable flags in parallel tests; verified with 100 shuffled race-detector runs.

Things I’ve built

A dependency health scanner for Go, npm, and PyPI. Combines vulnerability, deprecation, and release-staleness signals into a score you can gate CI on.

Bounded concurrency, deduplicated registry requests, and dependency tracing. Incomplete scans surface warnings instead of silently looking successful.

An embedded key-value store in Go, built with the standard library: an in-memory index backed by an append-only log.

An exercise in making durability concrete—when to sync, what survives a crash, and when a failed write should stop the database accepting more writes.

A pipeline that tests whether database backups actually restore. Led a five-person capstone team.

Dump → optionally corrupt → restore in an isolated container → compare data → report.

Built with Python, PostgreSQL, Airflow, and Docker Compose, with restore telemetry in a Next.js dashboard.


Tools I work with: Go, Python, Java, SQL · Linux, Docker, PostgreSQL, GitHub Actions, Airflow

Learning by building small systems, reading existing ones, and following failures across the boundaries.

Pinned Loading

  1. Resume Resume Public

  2. DistributedTaskManager DistributedTaskManager Public

    Python

  3. TinyDiff TinyDiff Public

    This is an AutoGrad engine inspired by MicroGrad by Andrej Karpathy.

    Python

  4. siaxvii/WatchListed siaxvii/WatchListed Public

    Your Personalized TV Show Recommendation App

    TypeScript 5 1