Profile
🪐 Science fiction put a stubborn question in my head early: what would it take for autonomous systems to feel trustworthy, not merely impressive? As modern AI tooling matured, that turned into quieter questions about representation, scale, and what very large models still miss on tasks humans treat as second nature.
⚖️ I try to hold research and engineering at equal weight. Shipping teaches you what breaks under load and regulation; research teaches you which abstractions are lying and where measurement still matters.
🎓 I completed a B.Tech. in Computer Science at KIIT, Bhubaneswar, with sustained work on cross-domain AI and ML alongside faculty, implementations from first principles, and attention to conferences such as ICDCIT and ICDAM. A LinkedIn internship opened the aperture on how large platforms stay reliable at petabyte scale, which pulled me toward software reliability and statistically serious coursework. I cleared GATE on the first attempt and now study Software Reliability (QRE) at IIT Kharagpur, with coursework spanning risk analysis, reliability prediction, distributed real-time systems, and adjacent statistics so those tools carry over to modern AI systems.
📄 I published my first technical paper while still in school and kept publishing through undergraduate research: a game-theoretic treatment of immersive virtual environments (MDPI Mathematics), follow-on cryptography and identity work including an ICDCIT book chapter from Code for GovTech (second place, PIC 2023), RSA structure studies (IEEE CISCT), an iFlow data-pipeline book chapter from ICDAM 2023, and CoMSeC on combined NLP embeddings for service classification (IEEE). The threads began in security and game theory and drifted naturally toward agents, evaluation, and data-heavy systems. Corpus (venues & dates) · C4GT full page →
📈 A parallel hobby in markets led me to WorldQuant BRAIN (Gold), which sharpened questions about latency, inference cost, and when smaller models or different architectures are the honest answer.
🤝 Moving from a quieter high school into collaborative engineering meant learning to reach out deliberately, on campus and online. The lesson stuck: most durable work is layered on other people’s foundations, so I bias toward clear collaboration and early alignment on constraints.
✉️ If you are building in research, reliability, aerospace, GovTech, or early products and want a precise thread, send a note. I enjoy building scalable systems, automating workflows, and solving problems where the measurement plan is as important as the demo.
Build
TypeScript · Python · Go · Rust · C++ · React · Node.js · Distributed systems
Ship
Kubernetes · DevOps · Observability · PostgreSQL · Redis · Cloud architecture
Decide
Finance · Product sense · Design thinking · Research · Scientific writing
Move
AI / ML · Blockchain & Web3 · E-government · Algorithmic trading · Open source
Words from collaborators, lightly trimmed for this canvas. Full context lives on LinkedIn. Use the arrows or dots to move between voices.
“Ansh is one of the most amazing engineers I have met. He is passionate about a wide variety of things… his knowledge spans a breadth of topics… he loves deep diving into things… a reliable person that could be trusted with juggling multiple work-streams.”