I'm Vanshaj Kerni, a physics-trained researcher with interests in astrophysics, observations, scientific ML, Agentic AI, differential equations and technology. I hold Master's in Physics (minor Astronomy) from Indian Institute of Technology Roorkee honoured with Department Gold Medal. Published two peer-reviewed papers, presented work in 4+ conferences, associate member of the American Physical Society, Astronomical Society of India. Earlier associate member of the Indian Pulsar Timing Array and the Indian Physics Association.
I currently work as a Junior Research Scientist at Arithmic Labs, where I build research-driven products—from ZK proof systems to LLM-based agents and SDKs for real-world use. Looking to go into the quantum industry with academic, software and research orientation. Aiming to merge astrophysics and quantum computing research.
- 📚 Academic Research: Authored two peer-reviewed papers across astronomy and applied mathematics.
- 🚀 Leadership: Former Additional Secretary of Physics and Astronomy Club, spearheaded astronomy outreach activities, headed organising scientific lectures from renowned physicists.
- 🌍 Global Projects: Worked with international teams across technical writing, core astronomy, and computational physics.
- 🧠 Gen AI Product Builder: Building AI-driven products for the general public and the astronomy community. Checkout: (Cite-Right)[cite-right.vercel.app]
- 🧬 ZK Proof System Optimisation – Built PoCs on zkVMs (Jolt) and explored performance tuning in proof generation.
- 🧠 LLM Agentic Systems – Built auto-prompting flows and agents using LangChain, OpenAI, and Gemini APIs.
- 🌐 Cross-functional Projects – Participated in projects combining astronomy, numerical physics, and core dev writing.
- 📖 Scientific Writing – Published work in astronomy and mathematics journals; co-authored concept papers and BRDs.
- Quantum Algorithm Usage in astronomical fluids
- A2A protocol and its real-world implementation
- LLM usage with RAG across verticals
- Integrating LLM agents with blockchain systems
- Architecting faster agent pipelines for low-latency systems