Systems-level thinker. I build tools that sit between C++ and Python — fast engines, search systems, and things that shouldn't exist but do.
I write performance-critical software and ship it to PyPI. My work spans native C++ acceleration layers, retrieval-augmented generation pipelines, and the occasional programming language built for fun. Currently contributing to open source through GSSoC '26 and maintaining projects used by developers across the ecosystem.
📍 Jaipur, India · 🎓 3rd Year CS Undergrad · 🚀 Building in public since 2024
~ $ whoami
anish-raj
~ $ cat current_focus.log
> Optimizing C++ memory arenas for zero-copy Python data bridges
> Building hybrid dense/sparse search systems without the vector-DB bloat
> Writing lexers and parsers from scratch for custom languages
| Languages |
|
| Core Stack |
|
| Infra & Tools |
|
| AI / ML |
|
|
C++ accelerated CSV processing engine for Python High-performance data preprocessing library with a native C++ backend exposed through pybind11. Designed to supercharge pandas workflows — handles type inference, null detection, and column transforms at near-native speed.
|
FAISS + BM25 + GPT-4o retrieval pipeline Production-grade retrieval-augmented generation system combining dense and sparse search strategies. Achieves 95% Hit Rate@5 with intelligent re-ranking and context-aware generation.
|
|
A programming language in Bhojpuri Open-source interpreted language built for the Bhojpuri-speaking world. Custom lexer, parser, and runtime — because every language deserves a programming language.
|
Offline-first DSA practice CLI Terminal-based tool for structured DSA practice. Metadata-first architecture with SQLite persistence, Rich-powered UI, and curated problem sets — no internet required.
|
