RAHUL GUPTA · BACKEND DEVELOPER

I build backend systems that hold up beyond the demo.

Distributed systems, real-time infrastructure, event-driven pipelines, and AI applications for products that need more than a prototype.

View my work →Let's talk →

Currently building production systems at CognoVerse

PRODUCTION BACKENDSDISTRIBUTED SYSTEMSREAL-TIME APPLICATIONSAI / RAGEVENT-DRIVEN ARCHITECTURE

15K+

purchase orders processed monthly

2K+

concurrent users supported

100ms

production API latency improvement

10K

concurrent jobs load-tested

Selected Work

Selected Engineering Work

Systems built around real constraints — reliability, concurrency, asynchronous processing, data, and AI.

DISTRIBUTED SYSTEMS

Distributed Fault-Tolerant Task Queue

A Redis + Lua distributed task queue with at-least-once delivery, heartbeat lease recovery, an independent Janitor reaper, Dead Letter Queue routing, and a React observability dashboard.

Node.jsTypeScriptRedisLua+5
View case study →
DISTRIBUTED SYSTEMS

Distributed Tiered Rate Limiter

A Redis + Lua rate limiting system with three interchangeable algorithms, per-API-key tiers, Prometheus metrics, k6 load tests, and a live React dashboard.

Node.jsExpressRedisLua+7
View case study →
EVENT-DRIVEN / MEDIA

StreamHub — Distributed Video Pipeline

An event-driven video platform: presigned direct-to-storage uploads, BullMQ + ffmpeg HLS adaptive-bitrate transcoding, and a custom hls.js player — without the API ever buffering video bytes.

FastifyReactTypeScriptBullMQ+7
View case study →
REAL-TIME SYSTEMS

LogStream

An event-driven log-processing platform using FastAPI, Redis Streams and asynchronous workers for high-throughput log ingestion.

FastAPIPythonRedis StreamsPostgreSQL+3
View case study →
AI / SEARCH

Research Paper RAG System

A CPU-optimised research-paper RAG API with FastEmbed, BM25 + RRF fusion, cross-encoder reranking, shared-vector multi-user access, and conversational query rewriting.

FastAPIPythonQdrantPostgreSQL+6
View case study →
AI / RAG

AI Medical Report Summarizer

A RAG system that analyzes and summarizes complex medical reports with FastAPI, LangChain, Qdrant, and a TypeScript client — fully Dockerized.

PythonFastAPILangChainQdrant+2
View case study →
REAL-TIME / COLLAB

DrawApp — Collaborative Whiteboard

A real-time collaborative whiteboard with Socket.IO, Redis Pub/Sub scaling, Redis job queues for persistence and email, NextAuth, and a Next.js + Tailwind UI.

Next.jsSocket.IORedisPostgreSQL+3
View case study →
REAL-TIME / BACKEND

Real-Time Chess App

A multiplayer chess backend with dedicated WebSocket infrastructure, durable game history, and live state management.

Node.jsTypeScriptPostgreSQLRedis+5
View case study →

Experience

Production Experience

CognoVerse

Back End Developer

May 2026 – Present

  • —Reworked the pagination layer on a core Node.js REST API, cutting average response latency by 100ms in production for the live mobile app.
  • —Owned end-to-end delivery of Common Carrier Phase 2 — built both the Node.js backend and the React web frontend solo, shipping a production-ready module across the full stack.
  • —Built a one-click ClickUp ↔ Excel sync automation for an Indian client, removing 3–6 hours of daily manual spreadsheet reconciliation.
  • —Architected and solo-built Olive, a full-stack deliverable-tracking and workflow-orchestration platform (Fastify on Bun, TypeScript, PostgreSQL, Redis Streams) — first commit to production in under 2 weeks, still the sole engineer owning it in production.
  • —Designed an event-driven pipeline across 4 decoupled workers for cloud-folder provisioning, email notifications, automated deadline escalation, and file-submission detection.
  • —Built a dual-cloud storage abstraction with live Google Drive and Microsoft OneDrive (Graph API) backends behind a single interface.
  • —Designed tokenless magic-link workflows for external stakeholders alongside JWT-based role/department access for internal users — two coexisting auth models.
  • —Shipped a hand-built, zoomable Gantt-chart visualization in React and an immutable audit-log system tracking 30+ distinct action types.

CognoVerse

SDE Intern

May 2025 – May 2026

  • —Developed an AI agent to automate sales workflows with a multi-agent LLM pipeline (LangGraph, FastAPI), Gmail/HubSpot integrations, Qdrant, OAuth token refresh, and real-time SSE streaming.
  • —Built a document-intelligence pipeline that ingests invoices from email (PDF, any structure, multipage), extracts metadata, and matches customer IDs and product codes from master data — processing 15,000+ purchase orders monthly.
  • —Worked with Panasonic on invoice ingestion requiring heavy preprocessing before structured storage.
  • —Built a medical-firm order ingestion system from master email across .txt, .docx, .pdf, and .xlsx, storing structured rows in DB and Excel with embedding-based master-data lookup for customer IDs.
  • —Built asynchronous fault-tolerant workflows using Celery, Redis, and PostgreSQL.

Beamstacks

Associate Application Developer Intern

September 2024 – March 2025

  • —Built an internal employee tracking app for weekly working hours, deployed on Azure with Azure Blob Storage for attachments.
  • —Delivered a Proof of Concept for Panasonic on RAG + AI that converted into a full project; also built a POC extracting structured text and images from PDFs into HTML.
  • —Developed the backend of a full-stack sports application with real-time score updates, clock synchronization, multi-game support, and end-to-end authentication — achieving sub-100ms state sync for 2,000+ concurrent users.
  • —Scaled the sports app with Redis queues/schedulers and Redis Pub/Sub so WebSockets fan out across multiple instances under load; deployed to Azure Ubuntu and AWS EC2.
  • —Built FastAPI conversational AI services using Qdrant, Redis caching, and LLM APIs.

Sales Assist

SDE Intern

April 2024 – June 2024

  • —Integrated Phantom on the backend and designed systems to reduce latency with well-optimized APIs.
  • —Created backend schedulers that run without impacting API performance.
  • —Helped automate LinkedIn activity workflows and processed large CSV datasets with pandas.

Sales Assist

React Developer

January 2024 – April 2024

  • —Built the application frontend with React Table and integrated dozens of APIs against Figma designs in Tailwind CSS.
  • —Implemented canvas-based infographic generation (text/hashtag separation) — a core VCMO feature — with social sharing and an emoji-safe caption parser.

Capabilities

What I Can Build

Backend Systems

  • REST APIs
  • Authentication
  • PostgreSQL
  • MongoDB
  • Redis
  • WebSockets
  • Background jobs
  • Microservices

Distributed Systems

  • Event-driven architecture
  • Message queues
  • Worker systems
  • Failure recovery
  • Dead Letter Queues
  • Rate limiting
  • Idempotency

AI Engineering

  • RAG
  • Vector search
  • Embeddings
  • Qdrant
  • LangChain
  • LangGraph
  • LlamaIndex
  • LLM integrations
  • Document intelligence

Infrastructure

  • Docker
  • AWS
  • Azure
  • Nginx
  • PM2
  • GitHub Actions
  • Linux
  • Prometheus
  • Load testing

Philosophy

How I approach engineering

Own the system

I like understanding a feature from schema and API design through workers, deployment, and observability.

Design for failure

Queues fail. Workers crash. Clients disconnect. Production systems need recovery paths.

Measure the system

Use benchmarks, load tests, logs and metrics rather than guessing about performance.

Stack

Technical Stack

Languages

TypeScriptJavaScriptPythonSQLLua

Backend

Node.jsExpressFastifyBunFastAPIRESTWebSocketsSocket.IOCeleryBullMQ

Data

PostgreSQLMongoDBRedisQdrantPrismaMinIO / S3

AI

LangChainLangGraphLlamaIndexRAGFastEmbedEmbeddings

Infrastructure

DockerAWSAzureGitHub ActionsNginxPM2PrometheusLinux

Testing

VitestPyTestk6

About

I like building the machinery behind the product.

I'm a backend developer focused on scalable systems, distributed architectures, and AI-powered applications. I enjoy problems around high-throughput queues, real-time communication, asynchronous processing, and data-intensive workflows — from schema and API design through workers, deployment, and observability.

At CognoVerse I own production backends for marketplace and deliverable-tracking platforms (Olive, Common Carrier). Before that I built real-time sports backends at Beamstacks, multi-agent sales automation and document-intelligence pipelines processing 15K+ POs monthly, and earlier full-stack work at Sales Assist.

BCA at DAV College, Panjab University (CGPA 8.4, 2nd in program; graduating April 2026) — but my identity is the systems I've shipped, not the degree line.

Class Rank #2

BCA program, DAV College, Panjab University

State Rank 10

Class X — HP Board, Nahan

300+

LeetCode DSA problems

Hackathon Runner-up

Chronicles Hackathon 2024

Building something that needs serious backend engineering?

If you're building a startup product, AI application, or backend-heavy system, I'd be happy to discuss the architecture, implementation, or a specific engineering problem you're facing.
Start a conversation →View GitHub →

Contact

Let's talk

The fastest way to reach me is email. I'm also on GitHub and LinkedIn.

rahul24012006@gmail.comgithub.com/RahulGIT24linkedin.com/in/rahu24