About
I'm Muhammad Waqar — a software engineer and AI training-data specialist with a competitive-programming background, currently finishing my B.S. in Computer Science at the University of Education, Lahore (class of 2027).
Now
- Contributing to CNCF open-source projects: OpenTelemetry Go compile-time instrumentation and Cognee.
- Preparing for LFX mentorships and building things with LangGraph and RAG pipelines.
- Competing in ICPC-style contests.
Work
TuringAI Evaluation Task Designer · Pod Lead
07/2025 – 03/2026 · Palo Alto (remote)
- Designed RLHF/SFT evaluation tasks and game-based procedural reasoning scenarios to benchmark frontier LLMs (Gemini, GPT-4, Claude), with Python validators that score outputs field by field and resist gaming.
- Led a pod of 8 trainers delivering C++ competitive-programming problem sets for fine-tuning pipelines across multiple clients.
- Co-reviewed 100+ problems for difficulty calibration against ICPC Regional / Codeforces Div. 1–2 standards, cross-referencing public archives to prevent training-data contamination.
Mama MarketplaceFull-Stack Engineer
01/2025 – 03/2025 · Switzerland (remote)
- Built full-stack features for a Swiss marketplace for second-hand baby, children, and maternity products (Next.js).
- Integrated Payrexx payment processing for secure checkout and transactions.
- Developed and maintained backend APIs supporting listings, users, and orders.
Headstarter AISoftware Engineering Fellow
07/2024 – 09/2024 · San Francisco (remote)
- Built and shipped 5 AI projects in 5 weeks (React, Next.js, Firebase) using agile sprints and CI/CD.
- Sharpened fundamentals with mock interviews and DSA / system-design practice.
Competitive programming
- Rank 30 at ICPC Asia Regional Contest
- #9 country-wide, #133 globally at CALICO Fall '25 (UC Berkeley)
- Bronze Brick at CALICO Fall '24 (UC Berkeley)
- 2nd Runner-Up at UE Code Quest
Toolbox
TypeScript, Next.js, React, Node.js · C/C++, Python, Rust · MongoDB, MySQL, Redis, Supabase · Docker, Kubernetes, AWS · LangChain, LangGraph, RAG
Contact
The fastest way to reach me is email. I'm also on GitHub and LinkedIn.