Full-stack & AI developer

I build AI-powered web apps, from the LLM workflow to the product around it.

Next.js, React, Node.js and FastAPI on the product side; LLM integration, tool-calling and RAG on the AI side. B.Tech ECE student at SRM University and Software Development Intern at ASA Industries.

About

I'm Utkarsh, an Electronics and Communication Engineering student at SRM University, specialising in VLSI and Digital Technology. I like building things end to end: the interface, the API, the data model and, more and more, the AI layer that ties them together.

At ASA Industries, an electric motor manufacturer, I replaced a 141-column Excel workflow with a production ERP and built a KPI system that computes department metrics straight from company data. That work taught me to check numbers before trusting them, a habit I bring to every project.

Right now I'm focused on agentic AI: LLMs that call tools, search the web and act on code. I'm looking for internships where I can build AI products that real people use.

Skills

AI & agentic AI

  • LLM integration: OpenAI API, Google Gemini, local models with Ollama, Puter.js multi-model access ContentSpy, Mood DJ, AI Algorithm Tutor
  • Tool-using agents: LLM calls with web search, model fallback chains, structured JSON output with repair ContentSpy
  • Code agents: a GitHub App that parses CI failures, proposes fixes and opens pull requests Devoploy (in progress)
  • RAG: OpenAI embeddings stored in Postgres with pgvector, retrieved for grounded answers Devoploy
  • Machine learning: scikit-learn, XGBoost, feature engineering, MLflow tracking Pre-Delinquency Risk

Languages

  • TypeScript
  • JavaScript
  • Python
  • Java
  • SQL
  • C++ (Arduino)
  • HTML/CSS

Frontend

  • React
  • Next.js
  • React Native (Expo)
  • Tailwind CSS
  • Framer Motion
  • shadcn/ui
  • Recharts
  • Vite

Backend

  • Node.js
  • Express
  • FastAPI
  • Next.js API routes
  • Prisma
  • BullMQ
  • REST APIs

Data

  • PostgreSQL
  • MongoDB
  • MySQL
  • Redis
  • pandas
  • NumPy
  • Excel modelling

Tools & cloud

  • Git & GitHub
  • GitHub Actions
  • Docker
  • Vercel
  • Render
  • Maven

IoT & hardware

  • ESP8266 (NodeMCU)
  • DHT sensors
  • PlatformIO
  • Arduino

Projects

ContentSpy AI dashboard with a competitor URL input

ContentSpy AI

Paste a competitor's URL and get an AI-written intelligence report: SEO keywords, content gaps, weaknesses and a market-entry plan. An LLM with web search does the research; reports are saved per user and export to PDF.

  • Next.js
  • TypeScript
  • Puter.js
  • Tailwind
Risk Intelligence Command Center dashboard with risk distribution and model metrics

Simulated data

Pre-Delinquency Risk Intelligence

An early-warning system that flags repayment stress before default: 16 behavioural signals from six months of banking activity, a Random Forest scorer, a FastAPI + MongoDB API and a React dashboard.

  • FastAPI
  • scikit-learn
  • MongoDB
  • React
Project Zenith landing page with a 3D globe and star field

Team project · ASTRALWEB'26

Project Zenith

An AI-powered digital observatory: live ISS position, satellites, planetary positions from NASA Horizons and an AI narrator, in an immersive 3D interface. My part: the Next.js frontend, 3D scenes, Cesium globe and deployment.

  • Next.js
  • Three.js
  • CesiumJS
  • FastAPI

In progress

Devoploy

A GitHub App agent for broken deployments: it clones the repo, parses CI and build failures, proposes a fix with an LLM and opens a pull request. Documentation search uses RAG over OpenAI embeddings in Postgres (pgvector). A worker on BullMQ and Redis runs the jobs.

  • TypeScript
  • Next.js
  • BullMQ
  • pgvector
  • OpenAI
AI Algorithm Tutor home screen with A* and alpha-beta modules

AI Algorithm Tutor

A React Native app that animates A* pathfinding and alpha-beta pruning step by step, with explanations from a local LLM served by Ollama, so it works offline.

  • React Native
  • Expo
  • Express
  • Ollama
Reverse DCF heatmap of value per share by revenue growth and cash-flow margin

Equity research

CG Power: Initiating Coverage

A full equity-research report with an Excel model (DCF, reverse DCF, peer comps, commodity sensitivity) and a Python script that rebuilds every valuation output and checks it against Excel.

  • Python
  • Excel
  • Matplotlib
  • Valuation

More projects

Experience

  1. May 2026 – Present

    Software Development Intern · ASA Industries

    Data, ERP & KPI systems · Uttar Pradesh

    • Built a production ERP for an electric motor manufacturer, replacing a 141-column Excel workflow across 5 departments (React, Node.js/Express, MongoDB Atlas).
    • Audited the legacy workbooks, rebuilt their calculations from raw inputs and reconciled every output; end-to-end tests across 6 modules showed zero inventory reconciliation drift.
    • Building a KPI system in Python that computes 83 department KPIs directly from company data, replacing hand-typed scores.
  2. Sep 2025 – Nov 2025

    Backend Developer Intern · Yuva

    • Built and documented REST APIs for e-commerce and blogging systems on Node.js and MySQL; load-tested with JMeter and cut API latency by 30%.
  3. Hackathons

    5 hackathons, second round in 2

    • Project Zenith for ASTRALWEB'26 with Team DO BRONXS; SwiftRelief, where I led backend and database integration.

Contact

Open to internships in AI and full-stack development. The fastest way to reach me is email.