Computer Engineering @ UWaterloo
I build software that holds up under pressure, and the tests that prove it.
Browser workspace for training & testing language models
— Launch training runs, compare checkpoints, and try models through streaming as-you-type autocomplete.
— Validated forms, live loss charts and logs, and a graceful stop that saves a resumable checkpoint.
— 127-test suite, zero skips: exact checkpoint replay, compiled-model correctness, evaluation behaviour.
Python / FastAPI / JavaScript / SSE
02Local-first personal knowledge base
— Captures notes, email, and browsing into searchable local storage with a plug-in system for new sources.
— Chrome extension with an offline IndexedDB outbox and backoff retries that survive network loss and restarts.
— Safe AI-agent edits: scoped views, transactional writes, idempotency keys, dry-run preview, audit log.
TypeScript / React / Bun / SQLite
03Tested simulation & analysis CLI
— Found, documented, and fixed three bugs in the original simulator.
— Added 8 Rust and 26 Python tests, linting, an end-to-end demo, and Windows checks to CI.
— Packaged an installable Python CLI for reproducible runs, plots, and model verification.
Rust / Python / GitHub Actions
04MCP server that lets an agent find the WaterlooWorks jobs that suit you best
TypeScript / MCP
05A toy language built on Brainfuck with real conveniences, running on a stack-based VM
Rust
06UW High Performance Computing onboarding challenge
C++
Software Engineering Intern (Forward-Deployed) · Python, SQL, AWS
— Adapted a B2B AI voice-agent platform to each client's workflow with Python integrations into their databases, CRMs, and scheduling tools.
— Worked directly with client engineers and stakeholders to review codebases, gather requirements, and plan integrations.
— Merged email, phone, and social interactions into unified customer profiles for the voice agents.
Mobile Development Intern · React Native, Laravel
— Halved app startup time by moving blocking initialization work to asynchronous operations.
— Built and documented React Native screens with designers, wired to APIs, databases, and server components.
— Wrote Laravel MVC backend code for apps handling national ID (CNIC) data; senior review found no security issues or bugs.
BASc, Computer Engineering · 2026 – 2031
IB Diploma · SAT 1600/1600 · 100% Merit Scholarship
languages: TypeScript, JavaScript, Python, SQL, Rust
applications: React, React Native, Laravel, FastAPI, Bun, SQLite, AWS, IndexedDB, Chrome extensions
quality & tools: Git, GitHub Actions, pytest, Cargo, Regression testing, CI, Root-cause analysis