Agniva Sengupta

/ɔɡ.niː.βɔ/ · noun · IIEST Shibpur '27

a full-stack engineer and an open source enthusiast.

a builder combining technical depth with creative design. primarily building for the web, occasionally diving into distributed systems and japanese culture.

status: shipping code + sipping coffee

Skills

JavaScriptJavaScript
TypeScriptTypeScript
ReactReact
Next.jsNext.js
Node.jsNode.js
TailwindTailwind
ExpressExpress
PostgreSQLPostgreSQL
MongoDBMongoDB
MongooseMongoose
PrismaPrisma
GitGit
GitHubGitHub
PostmanPostman
DockerDocker
RedisRedis
GolangGolang

The Build Journey

01

Distributed Analytics Engine

High-throughput event ingestion pipeline featuring load-leveling via Redis and asynchronous processing.

  • Achieved 1,626 req/sec throughput with <2ms latency.
  • Zero data loss reliability during traffic spikes.
  • Decoupled ingestion from processing using worker patterns.
  • Scalable microservices architecture containerized with Docker.
GolangRedisPostgreSQLDockerk6
02

PaperTrails

Production-grade blogging platform featuring decoupled traffic analytics and rich-text content creation.

  • Separate Go-based microservice for tracking article views.
  • Optimized NoSQL schema for complex user feeds.
  • Secure JWT authentication and real-time state management.
  • Interactive dashboard for content performance metrics.
MongoDBExpressReactNode.js
03

ClockedIn

Offline-first mobile attendance tracker featuring subject-wise scheduling and instant local persistence.

  • Fast reads/writes using MMKV storage engine.
  • Calendar-based class tracking and target setting.
  • Global state management for seamless onboarding.
  • Zero-latency offline availability for students.
React NativeExpoZustandMMKV

Current Explorations

01

Federated Learning

Researching privacy-preserving architectures for cardiovascular disease prediction. Focusing on FedAvg algorithms and decentralized model training.

02

Golang & Distributed Systems

Moving beyond basic syntax to understand concurrency patterns, microservices architecture, and high-throughput data processing.

Things About Me

I believe that great software is born from clarity.

My approach to engineering is shaped by a need for a clean slate—partly mandated by a dust allergy, and partly because a clear environment helps me think through complex problems.

When I'm not researching Federated Learning, you’ll find me analyzing Formula 1 race strategies, curating my Manga reading list, or brewing my daily fuel: Nescafe with a precise scoop of Amul Spray.

GET IN TOUCH

Connect with me on Linkedin or shoot an email.