Sumit Kumar Mahato
Focused on — Generative AI
Civil engineer turned AI engineer. Ex-Adani Solar and Healthark Insights, now building agentic AI systems.
Civil engineer turned AI engineer. I build production-oriented AI systems, from multi-agent workflows and RAG assistants to the machine learning models and dashboards behind real decisions.
At Healthark Insights I helped build Curie, an enterprise AI platform on FastAPI, LangGraph and PostgreSQL, with Redis caching, Docker and LangSmith observability. Before that, two years at Adani Solar commissioning a 2 GW TOPCon manufacturing line taught me how real systems get shipped.
I am currently pursuing a PGDM in AI & Data Science at Adani Institute of Digital Technology Management, and I am open to AI / ML engineering roles.
The short version, for —
AI & Data Science professional with 2+ years of industry experience, specialising in Generative AI, machine learning and scalable AI application development.
{ "name": "Sumit Kumar Mahato", "based_in": "Ahmedabad, India", "focus": "AI & Data Science", "stack": ["LangChain", "LangGraph", "RAG", "LLM Orchestration"], "studying": "PGDM in AI & Data Science", "status": "open_to_work"}
0+
Data & AI projects shipped
0.00
PGDM CGPA — out of 4.33
0GW
Solar plant build, Adani Solar
0+
Years of industry experience
Agentic & Generative AI
- LangChain
- LangGraph
- RAG
- LLM Orchestration
- Prompt Engineering
- Embeddings & Vector DBs
- Semantic Search
- Multi-Agent Workflows
- Tool Calling
- Human-in-the-Loop
- LLM Observability
Backend & Tools
- FastAPI
- REST APIs
- Docker
- Redis
- LangSmith
- Git & GitHub
- Postman
ML / DL & Statistics
- ANNs
- CNNs
- RNNs & LSTM
- Transformers
- Bagging & Boosting
- EDA
Programming & Data
- Python
- SQL
- PostgreSQL
- Power BI
- WorkIndia
- Developed Curie, an enterprise-grade AI platform using FastAPI, LangGraph, PostgreSQL and multiple API integrations to orchestrate intelligent, context-aware workflows.
- Built a scalable, observable AI backend with Redis caching, Docker containerisation and LangSmith for performance optimisation, streamlined deployment and end-to-end LLM monitoring.
- Engineered an AI-powered XLIFF translation pipeline using OpenAI GPT-4o, XML parsing and OCR-based graphics translation to automate Adobe FrameMaker localisation while preserving document structure.
(11)
Agentic AI, RAG assistants, machine learning models, computer vision and BI dashboards. Click any card to open its code on GitHub.
Keep scrolling
- 01IBMSQL and Relational DatabasesIBM
- 02PYPython for Data ScienceLearnbay
- 03IBMData Visualization with PythonIBM
- 04ADRevitAutodesk
Have a project in mind? I'd love to hear about it. Drop me a message and let's create something amazing.
Phone
+91 8757822221Based in
Ahmedabad, India