Open to opportunities

Rachit Roy

@rachitroy

Message

Full-stack developer building recommendation systems, cryptography-first apps, and ultra-fast databases.

India

Message

What I'm looking for

I’m looking for an entry-level full-stack or backend engineering role where I can build secure, high-performance systems—especially with recommendation/search, real-time features, and careful security/performance auditing—while growing into production-scale engineering.

I’m a full-stack developer with hands-on experience building production-grade applications across recommendation systems, applied cryptography, and database performance engineering. I built two full-stack applications end-to-end—architecting the React frontend and backend as one system—while keeping performance and security at the core of every decision.

In Buqs, I engineered a content-based recommendation engine with 1.6M+ similarity relationships across a 33,807-book catalog, delivering sub-millisecond to single-digit-millisecond query execution and a 99.5% database cache hit ratio. In Circle Chat, I implemented client-side end-to-end encryption using libsodium with Argon2id-based key wrapping and remediated security and efficiency issues after build—so I can ship real features without compromising trust. I’m seeking an entry-level full-stack or backend engineering role.

Experience

Work history, roles, and key accomplishments

Real-Time Encrypted Messaging

Circle Chat

Built an end-to-end encrypted real-time messaging application with a JWT-authenticated Socket.IO layer for live delivery, typing indicators, and presence tracking. Implemented client-side encryption using libsodium (X25519 + XSalsa20-Poly1305) with Argon2id-based key wrapping, and added secure encrypted image attachment handling via streaming to Cloudinary.

React Express MongoDB Socket.IO JWT Libsodium NaCl X25519 Argon2id XSalsa20 Poly1305 Client Side Security Optimistic UI Infinite Scroll Pagination Cloudinary Security Auditing N+1 Query Optimization (Aggregation Pipeline)

Full-Stack Book Discovery

Buqs

Architected and built a full-stack book discovery platform with a content-based recommendation engine over a large catalog, including feed/search/recommendation endpoints optimized for low-latency execution and high cache hit rates. Implemented recommendation and search using weighted cosine similarity and hybrid PostgreSQL search (full-text + trigram similarity), plus an async analytics pipeline.