Available for Opportunities · AI/ML Engineer

Hemil
Kaklotar

I build production AI systems — RAG pipelines, LLM integrations, and full-stack intelligent products that scale. 5+ years shipping real software across legal, health, and e-commerce.

View Projects Get in Touch
5+
Years Experience
10+
AI Systems in Prod
3
Clouds Deployed
9.0
CGPA · GTU
Scroll
// about me

Engineer by craft,
builder by obsession.

I'm an AI/ML Engineer and Full Stack Developer based in Surat, India. I specialize in turning complex AI research into production-ready systems — from RAG pipelines and fine-tuned LLMs to real-time video analysis and multi-agent frameworks.

My work lives at the intersection of language models, system design, and product engineering. I've shipped across legal-tech, healthcare, HR-tech, fitness, and e-commerce — always with a focus on reliability, latency, and measurable impact.

When I'm not building, I'm reading papers on alignment, contributing to open-source, or exploring the latest model releases from Anthropic, OpenAI, and Google DeepMind.

Core AI/ML
LangChain RAG Pipelines LLM Fine-tuning OpenAI GPT-4/4o Weaviate PyTorch Hugging Face RAGAS Evals LlamaIndex CrewAI
Backend & Infra
FastAPI Python Node.js Django Docker AWS GCP PostgreSQL Redis
Frontend
React.js Next.js TypeScript Tailwind CSS WebRTC
// experience

Where I've
built things.

Jul 2023 — Present
QONTAM
Self-Employed · Remote
AI/ML Engineer & Senior Full Stack Developer
  • Architected production RAG pipelines achieving <200ms retrieval latency on 100K+ doc corpora — cutting client query resolution time by 60%.
  • Built async ML inference serving >500 concurrent requests with <300ms P99 latency via FastAPI + Redis Queue.
  • Shipped 8+ production AI products across legal, healthcare, e-commerce, and fitness verticals.
  • Introduced RAGAS evaluation framework; iterated prompt strategies to improve LLM answer accuracy by 22%.
LangChainOpenAI GPT-4WeaviateFastAPINext.jsAWSGCPDocker
Jul 2022 — Jun 2023
iGeek
Surat, India
Senior Full Stack Developer
  • Integrated Azure AI Services to automate data extraction workflows, saving 3 FTE hrs/day and cutting processing time by 80%.
  • Mentored 4 engineers; code review culture lifted on-time delivery from 60% → 92%.
  • Deployed multi-cloud CI/CD reducing deployment cycle from days to <15 min.
Azure AINestJSReactTypeScriptCI/CDMongoDB
Jan 2020 — Dec 2021
Get Here IT Services
Mohali, India
Full Stack Developer
  • Delivered 6+ web applications; reusable React component library reduced dev time by 25%/project.
  • Python FastAPI microservices handling 10K+ daily API calls with zero critical incidents over 12 months.
React.jsFastAPINode.jsPostgreSQLMySQL
// selected projects

Things I've
shipped.

Legal AI · RAG
Intellifuse — AI Legal Document Intelligence
Production RAG pipeline ingesting 50K+ government legal docs. Achieved <150ms retrieval latency and 89% RAGAS faithfulness score via hybrid BM25 + dense search, adaptive chunking, and GPT-4 with citation-backed responses. Reduced user research time by 70%.
OpenAI GPT-4 Weaviate LangChain FastAPI RAGAS
Computer Vision · Fitness
OpenAir3D — AI Fitness Analytics
Pose estimation models (PyTorch + MediaPipe) with 94% accuracy across 12 workout types. Async inference at 24 FPS, <80ms/frame on AWS EC2.
PyTorch FastAPI Redis Docker AWS
HR Tech · Real-time AI
Tironem — AI Interview Platform
Azure AI video scoring for 200+ interviews/month. WebRTC with <100ms A/V sync. Cut recruiter screening time by 65%.
Azure AI WebRTC React TypeScript
Healthcare · Fine-tuned LLM
Xaviga — Mental Health AI Chatbot
Fine-tuned GPT-3.5 on 8K mental health samples — 38% relevance improvement vs base. 1,000+ active users, 99.5% uptime in 60 days.
OpenAI FT Node.js MongoDB React
SaaS · Conversational AI
Just Think AI — Conversational Platform
Multi-tenant AI SaaS. Context compression via Redis cut OpenAI costs by 30%. <500ms time-to-first-token with streaming responses.
OpenAI API Node.js Firebase AWS EC2
E-Commerce · Automation
Reeva.ai — AI E-Commerce Automation
GPT-4 SEO listing automation cut seller time from 20 min → <2 min/product. Multi-tenant PostgreSQL + Stripe for 3-tier subscription billing.
GPT-4 Next.js Django REST PostgreSQL
// contact

Let's build
something great.

Open to full-time roles, freelance AI/ML projects, and interesting collaborations. I respond within 24 hours.

hkaklotar17@gmail.com LinkedIn GitHub +91 87349 57975