# Prateek Mulye

> Senior Full-Stack Engineer, backend-heavy. 11+ years on distributed systems (Kafka, PostgreSQL at scale, financial-grade reliability), now bringing that same discipline to applied AI.

## Summary

I build distributed back-end systems: event-driven services on Apache Kafka, PostgreSQL tuned past 30M rows, clean recovery from bad input, and the contracts that hold up when traffic spikes. I learned that on financial and banking platforms where a mistake costs money. More recently I have brought the same discipline to applied AI: contributing to a manufacturing-AI pilot, and building agentic RAG and multi-agent systems in my own projects.

I am open to senior roles anywhere for the right team: international relocation, staying in the US, or remote. I am also learning German (A1-A2).

## Availability

- Open to senior roles anywhere for the right team: international relocation, staying in the US, or remote.
- 11+ years of professional IT experience.
- Email: prateekmulye@gmail.com
- LinkedIn: https://www.linkedin.com/in/prateekmulye
- GitHub: https://github.com/prateekmulye
- Hugging Face: https://huggingface.co/prateekmulye
- Site: https://prateekmulye.dev

## Experience

Roles are described by public label only; specifics are available on request.

- A manufacturing-AI initiative (2026 to present), Senior Full-Stack Engineer: contributing to a pilot that replaces spreadsheet-heavy plant work with AI you can trust. It answers from the actual procedure documents, and when it is not confident enough it says so and cites the source instead of guessing.
- A B2B data-intelligence platform (2021 to 2025), Senior Full-Stack Engineer: designed and built a data-cleanup service across 15M+ records, and scaled PostgreSQL past 30M rows so it stayed fast under heavy reporting; the slowest dashboards dropped from a wait to near-instant, more than 60% faster.
- A financial-grade transaction platform (2018 to 2021), Senior Backend Engineer: built the payment workflows where correctness mattered most, so a bad message is caught, retried, and replayed cleanly without ever double-processing; also built a service that held 12K+ requests per second in testing.
- A first online-banking platform (2013 to 2015), Software Engineer: built the secure login and anti-phishing layer in Java under real regulatory and performance constraints.

## Projects

- slipstream-f1-strategist (building, public): an event-driven race-strategy service in Java 21, Spring WebFlux, Kafka, PostgreSQL, Flyway, OpenTelemetry, Testcontainers. The public, NDA-free analogue of the financial-grade reliability work: bounded retry with backoff and idempotent upsert. No dead-letter-queue topic and no auto-replay. https://github.com/prateekmulye/slipstream-f1-strategist
- ChatFormula1 (live): an agentic RAG assistant for Formula 1, my most engineering-mature AI repo. A single-agent LangGraph router over a Pinecone vector store and Tavily web search, FastAPI backend, Pydantic-typed state, with production hygiene: auth, rate-limiting, tests, structured logging, caching, graceful degradation, and CI/CD. https://chatformula1.com
- FinResearch AI (live): a LangGraph multi-agent system I architected and deployed, a manager fanning out to specialized research agents with structured JSON outputs at every step. My submission to the SuperDataScience CP044 community project. https://huggingface.co/spaces/prateekmulye/FinResearchAI
- AegisHarness (building): a reference architecture that observes an AI agent's tool calls out-of-band, with an advisory policy gate and a tamper-evident audit. It informs and records, it does not decide what runs. Java, Spring Boot, Kafka.
- GlobalNomad AI (building): an early, pre-release build to help with the tax and visa side of moving between countries, a problem I am living through myself. Elixir, Phoenix, LangGraph.

## Stack

Java, Elixir/Phoenix, Python, TypeScript. Apache Kafka, RabbitMQ, AWS Kinesis, Oban Pro. PostgreSQL, Redis, DynamoDB. LangGraph, LangChain, RAG, Pinecone, Pydantic-typed state. AWS, Azure, Kubernetes, Terraform. OpenTelemetry, Datadog.