Headline

Backend Engineer (Node.js/TypeScript) shipping production ML & AI systems | XGBoost · LangChain · PostgreSQL | Open to SDE-1 & Founding Engineer roles | Remote-friendly | MIT-WPU ’26

About

I'm a backend-focused software engineer who ships production systems including production machine learning, not just prototypes. In my first two months as an Agentic AI Intern at Inttrvu.AI, I designed and deployed a Supplier Risk Intelligence System now in active use by the core engineering team and OEM audit teams across 16 suppliers and 5 OEMs, predicting 8 defect categories. It combines a 30+ feature XGBoost classifier (84% accuracy, F1 0.80, SMOTE-balanced) with a real-time Playwright scraping pipeline and an LLaMA 3.3/Groq-powered sentiment agent, surfaced through a multi-dimensional 0–100 health score on an interactive Streamlit dashboard. Outside work, I build and ship full systems end-to-end. Sportz is a real-time sports data platform (Node.js, Express 5, WebSockets, PostgreSQL, Drizzle ORM) that sustained 28,000+ requests with a 0.994 Apdex score and 140ms average response time under load hardened with bot-protection middleware and Zod-validated payloads across every endpoint. Global Uptime Monitor is a Turborepo monorepo (Next.js, Express, PostgreSQL/Prisma) with a real-time WebSocket dashboard and a Solana-based decentralized validator incentive model. I'm also a Hacktoberfest 2025 Golden Super Contributor top 10% of 98,000+ global participants with merged PRs into open-source projects including Kestra. Computer Science graduate from MIT World Peace University, Pune (CGPA 8.40/10, Class of 2026), available from August 2026. Based in Pune, open to backend, full-stack, AI/ML, and founding engineer roles remote-first, hybrid, or relocation across India and globally. If you're hiring engineers who can take a system from raw data to a production dashboard on their own, let's talk.

Location

Pune City

AI stack

aistudio

langchain

ollama

openai

Work experience

Agentic AI Intern at INTTRVU.AI, Mar 2026 - Present · 4 mos, Pune District, • Designed and deployed a production Supplier Risk Intelligence System covering 16 suppliers across 5 OEMs, predicting 8 defect categories and actively used by the core engineering team and OEM audit teams for daily risk assessment. • Engineered an end-to-end ML pipeline in Python 30+ engineered features, a multi-class XGBoost classifier (84% accuracy, F1 0.80), and SMOTE-based class-imbalance handling reducing critical defect false negatives and improving audit decision reliability. • Built a real-time data ingestion layer with a Playwright-based scraper across 16 supplier sources, paired with an LLaMA 3.3 (Groq)-powered risk agent for automated news sentiment analysis feeding directly into the scoring pipeline. • Architected a multi-dimensional Health Scoring system (Quality, Delivery, Finance, News) producing a 0–100 composite score with tiered risk labels, deployed via a 4-tab interactive Streamlit dashboard for OEM audit prioritization.

Open Source Contributor | Hacktoberfest 2025 at Hacktoberfest, Oct 2025 - Nov 2025 · 2 mos, Remote / Online, • Earned the Golden Super Contributor Badge top 10% of 98,000+ global participants through 7+ merged pull requests across open-source projects including Kestra (cloud workflow orchestration) and several DSA/UI repositories. • Collaborated with international maintainers using standard Git workflows and code review processes, contributing functional improvements, test coverage, and documentation to production-grade codebases

Education

Bachelor of Science - BSc, Computer Science, MIT World Peace University