# Pradhumn Gautam

> Pradhumn Gautam is a software engineer at QodexAI building backend infrastructure, reliable data systems, and AI products. Explore projects and engineering notes.

Canonical profile: https://www.pradhumn.com/

Software Engineer at QodexAI. Based in New Delhi, India.

B.Tech in Computer Science & AI from Maharaja Agrasen Institute of Technology, with a 9.1 CGPA.

## Experience

### QodexAI, Inc. — Software Engineer

Full-time, Jul 2025 – Present

- Architected fair-usage controls across a distributed worker fleet, enforcing per-user, per-team, and global quotas without compromising throughput.
- Built scalable ETL pipelines for third-party APIs with token-bucket rate limits, exponential backoff, and dead-letter handling.
- Designed asynchronous systems for real-time events and user automations, decoupling ingestion from downstream processing for better reliability.
- Improved PostgreSQL performance and consistency with atomic RPCs, targeted indexes, and query-plan analysis.
- Shipped an MCP server for internal agent tooling, Paddle subscriptions and feature gating, Slack and CRM integrations, and an operations dashboard.
- Led reliability work across a major platform migration and resolved authentication defects responsible for more than 5,000 daily errors.

### QodexAI, Inc. — Software Engineer

Internship, Jan 2025 – Jul 2025

- Built a Manifest V3 Chrome extension in TypeScript with background scheduling and resilient LinkedIn DOM workflows.
- Implemented shared SSO sessions across the web application and browser extension, along with rate limiting and analytics instrumentation.
- Developed profile and email enrichment, language detection, and LLM-based classification pipelines.
- Delivered OpenAI-powered product capabilities at the edge and supported production integrations end to end.

### Jeem Studio — Software Engineer

Internship, Jun 2024 – Oct 2024

- Integrated Replicate for image generation, enhancement, text-to-image, image-to-image, and multimodal workflows in Artext.
- Built Jeem Insights, a citation-backed RAG product with PDF ingestion, embeddings, vector search, and grounded retrieval.
- Designed asynchronous AI-processing APIs with FastAPI, Supabase, and PostgreSQL.
- Added Redis caching and fault-tolerant processing to improve latency and service resilience.

## Stack

- Languages: Python, TypeScript, JavaScript, SQL

- Backend: FastAPI, Node.js, Express, Supabase, MCP

- Data: PostgreSQL, Redis, MongoDB, SQLAlchemy

- Delivery: Docker, Docker Compose, uv, Alembic, GitHub Actions

- Frontend & AI: React, Next.js, OpenAI, LangChain, LlamaIndex

## Profiles and contact

- [GitHub](https://github.com/pradhumngautam)

- [LinkedIn](https://www.linkedin.com/in/pradhumngautam/)

- [X](https://x.com/iPradhumnGautam)

- [Hashnode](https://pradhumngautam.hashnode.dev)

- [Email](mailto:pradhumngautam0506@gmail.com)