Prajwal Murthy
Curiosity • Research • Learn • Build
About
I'm Prajwal Murthy, a senior software engineer at PayPal. I work across frontend, backend, and AI, and build open-source tools that I want to use myself. I spend a lot of my time on modern ML: reasoning models, AI agents, and making LLM inference fast and affordable, from smarter model routing to running open-weight models on your own hardware. I write about what I build and learn, and the systems behind everyday software.
Posts
Experience
Senior Software Engineer
PayPal
June 2025 - Present
Fullstack Engineer
Supplyframe Inc
June 2024 - June 2025
Senior Frontend Engineer
Certa.ai
March 2022 - November 2022
Software Engineer
Snapwiz Inc
August 2019 - March 2022
Skills
Languages
Frontend Stack
Backend Stack
Databases
Infrastructure
ML Tools
Open Source
Peekling
A tiny, private-by-default cursor companion for any website. Open-source, zero-dependency browser runtime with data-only character packs, no telemetry, and host-controlled interactions.
Oleafly
The local-first AI assisted research workspace for scientific writing & publishing. Research, Research, Write, Compile, Verify and Publish in LaTeX • Typst • Markdown • Git-native • Open Source • Desktop Performance
ThoughtFabric AI
Visual AI workspace built in Rust. Node-based context isolation, multi-model routing, and on-device RAG so knowledge synthesis stays private across different LLMs.
Education
Illinois Institute of Technology
Master of Computer Science
Visvesvaraya Technological University
Bachelors in Information Science and Engineering
Publications
Engine-Transfer-Bench: An Evidence-Based Benchmark for Document Compilation Engine Selection
arXiv (cs.SE) · Aug 19, 2026
arXiv:2608.18329
TeXFix-Bench: An Empirically Grounded Multi-Format Benchmark for LLM-Based Document Source Repair
arXiv (cs.SE) · Aug 6, 2026
arXiv:2608.07617
Patchwork: A Traffic Capture and Analysis Platform for Network Experiments on a Federated Testbed
Won the Best Community Contribution Award (ACM IMC 2025)
DOI: 10.1145//3730567.3764462
Using the FABRIC Testbed to teach a Graduate Computer Networking course
Published in Association for Computing Machinery (ACM)
DOI: 10.1145/3641554.3701923
Universal Semantic Web Assistant based on Seq-2-Seq Model and Natural Language Understanding
Published in IEEE Xplore and ICACC
DOI: 10.1109/ICACC48162.2019.8986173