Back to home

Solutions · AI & software, end to end

From your requirement to a system running at scale.

I design, build and deploy complete software: the interface, the AI, the backend, the cloud, and even the hardware when your idea needs it. One builder who owns the whole system, so nothing gets lost between teams.

Production healthcare SaaS · AWS deployments · Government patent · National hackathon wins

Sre VarshanApplied AI & GenAI Engineer Ships to AWS

End to end

Every layer of your system, built by one person who sees all of it.

Most projects break at the seams between teams. I build each layer myself, so the device, the data, the AI and the screen your users touch are designed to work together.

  1. 01

    Interface

    Web apps, dashboards and landing pages people enjoy using

  2. 02

    AI & intelligence

    LLMs, agents, OCR, vision and models that solve the actual problem

  3. 03

    Backend & data

    APIs, business logic, auth and databases that stay correct under load

  4. 04

    Cloud & delivery

    Deployed, monitored and ready to scale from day one

  5. 05

    Edge & hardware

    Sensors and devices that talk to the cloud reliably

What I build

Pick a starting point. I'll take it the whole way.

Shipped, not slideware

Real systems I have built and deployed.

Production · Healthcare SaaS

Gen AI inside a live credentialing platform

Seven months on a US healthcare product in daily use: a Text to SQL agent over PostgreSQL on AWS RDS, an OCR pipeline that autofills forms, deadline tracking for 60+ documents per provider, and Entra ID single sign on.

  • FastAPI
  • PostgreSQL · RDS
  • Qwen2.5 VL
  • Azure Entra ID
  • AWS
Read the experience

Edge to cloud · 4 days

Algae air purifier, device to dashboard

A working unit with sensors and a blower, streaming readings over LoRa and WiFi into API Gateway, Lambda and DynamoDB, with Cognito sign in, a FastAPI backend on RDS and a React dashboard on Amplify.

  • ESP32
  • API Gateway
  • Lambda
  • DynamoDB
  • Cognito
  • Amplify
See the project

Open source · Developer tool

TextLens, a reusable OCR framework

A Python framework that makes local document OCR simple to set up and reuse: hardware aware model selection, batch processing, caching and FastAPI serving, packaged so other developers can build on it.

  • Python
  • PyTorch
  • CUDA
  • FastAPI
  • Docker
See the project

How we work

Five steps from first call to live system.

  1. Discover

    We talk through your goals, users and constraints until the problem is clear.

    You get: A written scope you can hold me to

  2. Plan

    I design the system: architecture, data, integrations and a delivery plan.

    You get: Architecture and timeline

  3. Build

    Working software in short cycles, so you see real progress every week.

    You get: Weekly demos you can click

  4. Test

    Every flow checked end to end, including the edge cases and failure paths.

    You get: A tested, documented build

  5. Deliver

    Deployed to the cloud, handed over cleanly, and supported after launch.

    You get: Live system plus ongoing support

Working with me

What you can count on.

Portfolio template

Like this website? Get the whole thing for your own portfolio.

The full codebase behind this site, with placeholders where my content goes. Drop in your work and launch.

  • The complete Next.js + TypeScript source of this portfolio
  • Every section, animation and the pixel sky design system
  • Placeholder content and images, ready for your own
  • A step by step SKILL.md guide to customise and deploy it

₹399

One time payment

  1. 1
    Complete purchaseSecure checkout with Razorpay
  2. 2
    Get the codebaseYou get access to the entire codebase template
Checkout coming soon

Have a requirement? Let's turn it into a working system.

Tell me what you need. You get a clear plan for how to build it, whether or not we work together.