Hi, I am Rudraksh Agarwal!
- 01 Startup guy.
- 02 Worked in startups — lean teams, fewer than 10 engineers.
- 03 Tried building a startup of my own back in 2023: an agentic AI product, before the term "agent" had even been coined.
- 04 I don't believe in work-life balance. Work hours aren't over till the work is complete — or you hit the Claude 5-hour token limit.
- 05 I understand software, product, and a startup’s vision.
- 06 I believe I have high customer empathy.
- 07 I run hiring groups for SWE, ML, Data, Product, Design, and GTM. DM me if you want in.
- 08 I keep learning about agentic, backend, and distributed systems.
- Python / Node.js / TypeScript
- FastAPI / Flask / Temporal
- LangChain / LangGraph
- GCP / AWS / Docker
- PostgreSQL / PgVector / Redis
- MS SQL Server / Firestore / Firebase
- RabbitMQ / MQTT / Celery
- Prometheus / Grafana
Applied AI & Backend Systems
Open to Backend / Founding / AI Engineering roles
Jersey City, NJ · Open to relocate
Professional Experience
2019 — 2026Software Engineer · Faster Outcomes
Drove end-to-end development (frontend and backend) of a new healthcare platform built from scratch, partnering directly with PMs and customers to scope and ship features for wellness clinics. Built a ReAct-style agentic workflow that generates patient assessments via MCP tools over patient records, and a Temporal-based sync service ingesting files from third-party legal/EHR systems — 10K+ files synced across 100+ active cases.
Senior Software Engineer · Wipro Linecraft AI
Led a team of 4 engineers across the product's core services. Built a real-time, event-driven loss-detection system ingesting multi-logger streams every 500ms (cutting manual monitoring 50%), re-architected a core ETL pipeline in Python with Luigi for 6x faster runs, and owned the backend API layer powering the Manufacturing Analytics suite. Stood up Prometheus/Grafana monitoring that cut outage detection from hours to under a minute, and containerized 5 services into Docker.
Co-Founder · Light AI Labs
Conceptualized an Agentic product to reduce context switching in workspace in a Pre-Agentic Era.
Global Remote Mentorship Intern · IBM
Finetuned BERT on a technical corpus using MLM, NSP, and Sequential Question Generation. Built a PyTorch language model combining a multilayered BERT-based span selection model with an Attention-over-Attention architecture, trained on a GPU cluster — achieving an F1 score of 57.7 on the TechQA dev set.
ML Intern · My Next Film
Converted text stories into movie scripts using NLP. Worked with Coreference Resolution, POS Tagging, Dependency Parsing, and Named Entity Recognition.
Selected Projects
VIEW ALL GITHUBAgentic Dataset Curator
Agentic tool to clean raw datasets and chat with your data. Supports multi-step planning for complex data operations with natural language queries via LangGraph.
Agentic Job Application Drafter
An agentic workflow that polls RSS job feeds and drafts a tailored application per listing — using a Browser Use agent to extract form fields and an LLM to attach values and auto-fill applications.
Academic Foundations
2016 — 2025SUNY Buffalo
2024 — 2025Open to interesting problems.
Focused on agentic systems, distributed systems, and full-stack products. Always happy to connect.