Prateek Lachwani.

AI & DevOps Engineer

What I do

I build
AI-driven
ecosystems.

My stack

AI · DevOps
& Full Stack.

By the numbers

4+Years of Grinding
10+Projects Delivered
Stage Finishes
1+Year On-Field Experience

/ Selected Works

WHERE IDEAS BUILT TO LAST PURE OBSESSION

BECOME REALITY. SHIPPED LIVE. MADE REAL.

SCROLL TO WITNESS. SEE FOR YOURSELF. SCROLL DEEPER.

Each project, a proof of obsession

Selected Works
Health
LLM
Finance
Gaming
Ayurveda
AI
Agents
AI
Narrative
RAG

fin

/ About

AI & DEVOPS ENGINEERBUILDING intelligent SYSTEMSTHAT scale.

Who I am

Passionate about blending AI & infrastructure, I craft systems that push the boundaries of what software can do — from self-healing DevOps pipelines to medical-grade AI platforms.

My work ranges from LLM orchestration and RAG pipelines to cloud-native Kubernetes deployments, always aiming to create reliable, scalable, and beautifully engineered digital ecosystems for ambitious products.

4+Years Grinding
10+Projects Delivered
Stage Finishes

Stack & Skills

AI / ML

  • LangChain
  • LlamaIndex
  • OpenAI
  • Claude
  • RAG Pipelines

Backend

  • Python
  • FastAPI
  • Node.js
  • PostgreSQL
  • Redis

DevOps

  • Docker
  • Kubernetes
  • GitHub Actions
  • Terraform
  • AWS

Frontend

  • Next.js
  • React
  • TypeScript
  • Framer Motion

Experience

May 2025 – Aug 2025

AI/Data Intern — Azure OpenAI & Infra

Jodhpur, Rajasthan

  • Built an ETL pipeline using Azure Document Intelligence & OpenAI to scrape Hindi newspaper archives.
  • Transformed data via fine-tuned GPT models, exporting structured JSON with 90% accuracy.
Blessed Infotech
May 2025 – July 2025

DevOps Intern

Rajasthan

  • Deployed Dockerized insurance app on Azure VMs — 70% faster deployments, 99.9% uptime.
  • Built CI/CD pipelines via Azure DevOps; cut deployment errors 50%, release cycles 40% faster.
  • Orchestrated containers with Docker & Kubernetes (AKS), reducing provisioning time by 60%.
Celebal Technologies
2024 – Now

AI Engineer & Co-Founder

AI Health Tech

  • Building a medical-grade AI platform that turns health chaos into high-precision intelligence.
  • Architecting LLM pipelines, RAG systems, and cloud-native infrastructure end-to-end.
Tabcura
May 2024 – July 2024

Software Engineer Intern

Bengaluru, Karnataka

  • Validated AlgoQA (v4) on Playwright TypeScript, increasing verified actions by 30%.
  • Boosted test coverage 40% using Java Selenium, Appium & Gherkin for SaaS automation.
  • Identified 116+ non-functional actions on GitHub, improving product stability.
AlgoShack Technologies

/ Feedback

WHAT PEOPLEARE saying.

Arjun M.

Prateek built our entire AI pipeline in weeks. The LangChain + RAG setup he architected handles 10k queries a day without breaking a sweat.

Arjun M., CTO at MedStartup

Sarah K.

Best DevOps hire we've had. He set up our entire CI/CD on Azure DevOps, cut deployment time by 60%, and documented everything perfectly.

Sarah K., Engineering Lead at CloudCo

Rohan S.

Prateek doesn't just write code — he thinks in systems. Our Kubernetes cluster went from chaos to 99.9% uptime under his watch.

Rohan S., VP of Engineering

Meera P.

The AI document intelligence pipeline he built for us was insane. 90% accuracy on Hindi newspaper OCR using fine-tuned GPT — I didn't think it was possible.

Meera P., Head of Data at Infotech

James T.

He shipped a full-stack MERN app with Docker + AKS in under 3 weeks. Would hire again without a second thought.

James T., Product Manager

Pooja R.

Prateek is the rare engineer who can go from LLM orchestration to infrastructure. He's the definition of full-stack AI engineering.

Pooja R., Co-Founder at HealthTech

Alex D.

Our test coverage jumped from 40% to 80% after Prateek set up our Playwright + Selenium suite. Solid, reliable, and actually clean code.

Alex D., QA Lead at SaaS Company

Vikram N.

He onboarded faster than anyone I've worked with and immediately started delivering. A true self-starter who knows AI infrastructure inside out.

Vikram N., CTO at Fintech Startup

Nina C.

Tabcura's architecture under the hood is genuinely impressive. Prateek designed it from scratch and it scales beautifully.

Nina C., Healthcare Investor

Siddharth A.

The MCP orchestrator he built handles multi-agent LLM workflows better than tools 10x the cost. Pure genius execution.

Siddharth A., AI Researcher