RAJ PRATAP SINGH GURJAR
Software engineer · AI systems · researcher
I am a Mathematics and Computing student at IIT Patna who enjoys working where software systems, machine learning, and careful reasoning meet. I build products end to end, but I am especially interested in inference speed, evaluation, and the design choices that make a system dependable. My recent work spans AI tools, database internals, clinical software, and research in legal reasoning.
Incoming
Software Engineering Intern · Google
Summer 2027
An upcoming role, not current employment. I will be joining Google for a software engineering internship in Summer 2027.
Experience
Research Intern
Research work with David Windridge on neurosymbolic and legal reasoning, structured inference, and evaluation.
Research Operations - Coding Expert (Contract)
Contributed to coding and evaluation workflows for improving model quality.
Software Engineer Intern
Built the Patient Pathway portal for ENT doctors from the ground up.
Education
2024 - 2028B.Tech in Mathematics and Computing
Indian Institute of Technology, Patna
Relevant Coursework: Machine Learning, Data Science, Probability & Statistics, Linear Algebra, Discrete Mathematics, Optimisation, Algorithms, Database Systems, Operating Systems, Numerical Methods
2022 - 2024Senior Secondary (12th)
Bhaskar Academy, Indore
Relevant Coursework: Physics, Chemistry, Mathematics
2012 - 2022Secondary School (10th)
Delhi Public School, Gwalior
Relevant Coursework: Mathematics, Science, Computer Applications
Featured Projects
disql
A lightweight SQL engine and query playground built to understand databases from the inside out, including parsing, execution, indexing, and the practical trade-offs behind a database interface.
Tewtr
Multi agent AI pipeline converting PDFs to structured learning content. Extreme fast inference cuts processing from 30min to 12s. Token throughput exceeding 600+ tokens/sec which makes it almost 5× faster than frontier AI.
DBMS Query Simulator
Custom DBMS engine supporting SQL-like parsing, execution, and indexing with B-Tree and Hash structures. Features Two-Phase Locking (2PL) and a React.js interactive visualizer.
Tools I use
PyTorch
C++
Python
TensorFlow
DockerCurrent Focus
Inference systems
Efficient model serving and systems performance
Neurosymbolic AI
Combining learned models with explicit structure
Quantitative systems
Market microstructure, C++, and event-driven research
Open Source
Contributing to open source projects
Achievements
- —600+ algorithmic problems solved across Codeforces, CodeChef, and LeetCode.
- —Achieved top 0.3 percentile in JEE Mains 2024 (1.4M candidates) and top 3.2% in JEE Advanced 2024 (250K candidates).
- —Completed the Google Upskilling Launchpad Program (2025), a competitive initiative focused on advancing technical and professional skills through hands-on projects and mentorship.
- —Qualified for AIME 2024 (Ranked 32/1000+ nationally) and earned IOQM Merit Certificate (Top 1.2% of 250,000 candidates, 2022).
- —Codeforces – Pupil (1328), and Placed 1,812/45,000+ in the IICPC Codefest Prelims (Feb 2025).
- —Research Consultant, WorldQuant – Awarded Gold Certificate.
- —Among only two first-year teams selected from IIT Patna to represent at SIH 2024.
Get in Touch
Feel free to reach out for collaborations, opportunities, or just to say hi!