Oluwatobi Olajide
I like building from first principles. Give me a protocol, a messy spec, or a blank folder, and I will turn it into a working system that you can run, test, and understand. I enjoy the parts of software that touch real constraints: network timeouts, disk layouts, message formats. I try to write code that is simple to trace and easy to change later.
Last summer I worked as a Software Engineer intern at Microsoft. I learned how to ship inside a large codebase, review with intention, and land changes that other people depend on. That experience made me sharper about tradeoffs and clearer in how I communicate. I still take the same approach on my personal projects. Keep the loop tight. Measure early. Prove it works.
On GitHub I keep a mix of learning builds and tools. You can find me at @tobibytes. My profile links to my site and LinkedIn, and it highlights the kinds of repos I want you to see first.
Selected Projects
BitTorrent client in Rust
I built a BitTorrent client to learn the protocol end to end and to get more comfortable with Rust. The repo shows a focused, test-guided build with Rust as the main language and a small shell wrapper for running steps. You will see the usual project scaffolding, a src folder, a sample torrent for local testing, and a CodeCrafters config that keeps the progression honest. This project taught me a lot about message framing, parsing, and state management under network pressure.
Redis clone in Python
This is a simple Redis-style server built to practice protocol handling, request routing, and small performance decisions in Python. The code is mostly Python with a tiny bit of shell to drive execution. It lives in a clean directory layout with an app folder and a straightforward README. I like projects like this because they force clear interfaces and help me practice writing code that is both minimal and testable.
Talvra
Talvra is my learning management system project. It is a monorepo that brings a modern stack together so students can process course content and generate study aids with AI. The repo uses a pnpm workspace with a React frontend, and a services layer planned in Node and Python. For storage and messaging I set up Neon Postgres, Redis Streams, and Azure Blob Storage. Docs in the repo outline the task plan, contributor workflow, and a Canvas LMS integration path. The README also spells out path aliases, local infra with Docker Compose, and health checks so you can run each piece in isolation.
See more on my GitHub profile: github.com/tobibytes