Hi, I'm

Darsh Bothra

AI Engineer building production-ready AI systems.

ECE undergrad at NIT Surat. I work with GenAI, LLMs, AI agents, and backend systems, building applications that move from idea to deployment.

View ProjectsDownload Resume

About

I'm an Electronics & Communication Engineering undergrad at NIT Surat, focused on AI engineering and backend systems. I work with GenAI, LLMs, AI agents, FastAPI, and retrieval pipelines — shipping systems that run in production, not just notebooks. I care about inference speed, clean APIs, and products people can actually use.

Experience

  1. Jan 2026 – Jul 2026

    Indore, Madhya Pradesh

    AI Engineer Intern

    Quasi Intelligence

    • Built FastAPI + HTMX backend services for industrial AI applications with real-time inference.
    • Integrated Modbus industrial communication to connect backend services with PLCs/devices for machine data ingestion.
    • Trained and optimized RF-DETR and lightweight CNNs on 100K+ industrial images; deployed a TensorRT engine on Jetson Orin Nano with 30% lower CPU utilization.
    • Contributed to a GraphRAG document intelligence system using LangGraph, Qdrant, Ollama, and hybrid retrieval, improving retrieval accuracy by 10%.

Projects

Selected work in RAG systems, analytics agents, and AI products.

AI Doc QA
Async FastAPI RAG system for document ingestion and question answering over uploaded PDFs.
PythonFastAPIPostgreSQLQdrantOpenAIDockerJWT
  • Built an async FastAPI backend with PostgreSQL, JWT authentication, and document ingestion.
  • Implemented a RAG pipeline with OpenAI and Qdrant for document retrieval and QA.
  • Added GitHub Actions CI/CD with automated linting and tests on pushes and pull requests.
GitHub
Expense Analyzer
Expense analytics app with CSV/XLSX ingestion, tool-calling LLM agent, and production AWS deploy.
PythonFastAPINext.jsDockerAWS
  • Deployed on AWS EC2 with Docker, Nginx, HTTPS, and a custom domain.
  • Built a FastAPI backend with CSV/XLSX ingestion and a tool-calling LLM agent for reliable analytics.
  • Implemented latency, token, and cost tracking with a monitoring dashboard.
GitHubLive Demo
Mock Mate
Real-time AI mock interview platform with voice interaction and automated performance reports.
Next.jsVapiGemini APIFirebase
  • Built a real-time AI mock interview platform using Next.js, Vapi voice API, and Firebase.
  • Integrated Gemini API to evaluate interviews and generate automated performance reports.
  • Implemented secure user authentication with Firebase Authentication.
GitHubLive Demo

Skills

Languages

PythonJavaScriptC/C++HTMLCSS

AI / ML

PyTorchLangChainScikit-LearnOpenCVNumPyPandas

Frameworks

FastAPIExpress.jsReact.jsNext.js

Databases

PostgreSQLMongoDBQdrant

Tools

GitGitHub ActionsDockerAWS

Contact

Interested in building AI products or scalable backend systems? Let's connect.

EmailLinkedInGitHub