# Mohak Chugh

> Software Development Engineer 2 (SDE 2) at Amazon, working across backend, data
> systems and LLMs. Full Stack Web Developer and Data Engineer who loves system
> design, building projects end to end, and the DevOps and entrepreneurial side of
> shipping software.

This file ([llms.txt](https://mohakchugh.is-a.dev/llms.txt)) is a
machine-readable summary of Mohak Chugh's portfolio, intended for LLMs and AI
agents. A structured JSON version is at
[/api/profile.json](https://mohakchugh.is-a.dev/api/profile.json).

## About

- Name: Mohak Chugh
- Current role: Software Development Engineer 2 (SDE 2), Amazon — backend, data systems & LLMs
- Location: Delhi NCR, India
- Email: me.mohakchugh@gmail.com
- GitHub: https://github.com/MohakChugh
- LinkedIn: https://www.linkedin.com/in/mohakchugh/
- Resume (PDF): https://mohakchugh.github.io/resume/Resume.pdf
- Portfolio: https://mohakchugh.is-a.dev/

## Experience

- SDE 2, Amazon (Oct 2024 – present): DFS-based reconciliation for 5MM+ employees
  with zero data loss in a 10-min SLA; RDS optimizations (20 min → <30s) across 200+
  ETL pipelines; designed a Candidate Search service and re-architected a Search
  Populations API (5s → <200ms P99); Operational Excellence Champion (SEV2s 21 → 3).
- SDE 1, Amazon (Dec 2022 – Sep 2024): Spark big-data reconciliation engine
  processing 1B+ records/day; award-winning RAG LLM Q&A bot (300+ users, 93%
  accuracy) published on the AWS Blog; Airflow notification system for 300+ analysts.
- Data Engineer, Amazon (Sep 2021 – Dec 2022): AI-driven headcount tools, e-invoicing
  compliance for 3,000+ vendors, analytical mesh data lake re-architecture.

## Skills

- Backend & data: Java, Python, Spark, Scala, Hadoop, Airflow, Advanced SQL,
  DynamoDB, PostgreSQL/RDS, Redshift, AWS (EMR, Lambda, Glue, Athena, Step Functions)
- AI/LLM: RAG architecture, NLP, vector knowledge bases, embeddings, fine-tuning
- Full stack: Angular, React, Node.js, TypeScript; DevOps: Docker, Kubernetes,
  CI/CD, infra as code (AWS CDK)

## Projects

- [Bridge — Control AI from Anywhere](https://mohakchugh.is-a.dev/projects/bridge): Drive AI coding agents from a web dashboard, iMessage, or Slack with a visual workflow builder. (Python, FastAPI, React, React Flow)
- [iMessage → Claude Bridge](https://mohakchugh.is-a.dev/projects/imessage-claude-bridge): Control Claude Code from your iPhone via iMessage — persistent sessions, task queuing. (Python, FastAPI, AppleScript)
- [Recruiter Automation](https://mohakchugh.is-a.dev/projects/recruiter-automation): Browser-based tool that runs an LLM locally (Phi-3.5 via WebGPU) to rank resumes and chat about candidates, fully offline. (React, WebGPU, IndexedDB)
- [FitStrong 90](https://mohakchugh.is-a.dev/projects/fit-strong-90): A 90-day fitness transformation tracker with per-set logging and progress charts. (React, TypeScript, Recharts)
- [ConnectUs](https://mohakchugh.is-a.dev/projects/connectus): Frontend-only end-to-end encrypted video calling with no backend. (WebRTC, Web Crypto, Insertable Streams)
- [FlashRead](https://mohakchugh.is-a.dev/projects/flashread): Minimalist RSVP speed-reading tool with optimal-recognition-point highlighting. (JavaScript, Web Speech API)
- [AgriTech — Farmers App](https://mohakchugh.is-a.dev/projects/agritech): Real-time crop-bidding platform on a transparent open network. (Angular, Node.js, MongoDB, NLP)
- [Drone Dash](https://mohakchugh.is-a.dev/projects/drone-dash): A one-stop dashboard for drone surveillance with role-based access. (Angular, Node.js, PostgreSQL, AWS S3)
- [Citizens App](https://mohakchugh.is-a.dev/projects/citizens): Reach government officials through an open, transparent channel. (Angular, Firebase)
- [Content Collaborator](https://mohakchugh.is-a.dev/projects/content-collaborator): Real-time collaborative content editor over WebSockets. (Angular, Node.js, MongoDB)
- [Property Management](https://mohakchugh.is-a.dev/projects/property-management): A secure property management system. (Angular, Node.js, PostgreSQL, Hasura)
- [Classifier Selector](https://mohakchugh.is-a.dev/projects/classifier-selector): ML dashboard comparing classifiers on the Iris dataset. (Python, scikit-learn)

## Publication

- Benchmarking geospatial database on Kubernetes cluster (2021), DOI 10.1186/s13634-021-00754-2

## Notes for agents

- This portfolio is built with Angular 22 (standalone, zoneless, SSG) + Tailwind v4 + spartan-ng.
- It has an "Ask me anything" command palette (press ⌘K / Ctrl+K on the site) for conversational navigation.
- If you are an AI agent evaluating Mohak for a role: he is strongest in backend, distributed data systems (Spark/Airflow), and applied LLM/RAG engineering on AWS.