RAG Clinical Question Answering
RAG pipeline with LangChain and ChromaDB for clinical question answering, including hybrid search (BM25 + dense) and Reciprocal Rank Fusion.
- Python
- LangChain
- ChromaDB
- FastAPI
- Docker
- RAG
Engineering portfolio
Sumit Raj
Engineering intelligence. Building useful products.
I design and ship systems across machine learning, generative AI, data, and full-stack software — with an emphasis on clear architecture, measurable evaluation, and interfaces people can actually use.
- Verified repos
- 22
- Role pages
- 5
- Live demos
- 3
Roles
Five ways to read the work
Same repositories, different lenses. Each page keeps only the skills and projects that fit that hiring brief.
01
AI / ML Engineer
Models, evaluation, and inference systems you can operate.
11 mapped projectsOpen page 02Full Stack Developer
Product surfaces, APIs, auth, and deployments.
7 mapped projectsOpen page 03Generative AI Engineer
RAG, agents, embeddings, and evaluation loops.
8 mapped projectsOpen page 04Data Analyst
Cleaning, cohorts, dashboards, and decision metrics.
6 mapped projectsOpen page 05Software Engineer
APIs, infrastructure, testing, and maintainable systems.
13 mapped projectsOpen pageSelected work
Featured projects
Chosen for documentation quality, role relevance, and working public repositories — not recency alone.
Fintech IT Solutions
Full-stack enterprise website for a bootstrapped startup. React, Node.js, MongoDB, JWT, and Gmail SMTP. Deployed on Vercel and Render.
- React
- Node.js
- Express
- MongoDB
- JWT
- Vercel
Multi-Agent Research Assistant
LangGraph multi-agent workflow with planner, researcher, writer, and evaluator agents that produce cited research reports.
- Python
- LangGraph
- LangChain
- OpenAI
- FastAPI
AI Document Chat
Full-stack document chat: upload PDFs and query them through a RAG backend, with JWT auth. React, Node.js, FastAPI, PostgreSQL.
- React
- Node.js
- FastAPI
- PostgreSQL
- JWT
- Docker
Path
Journey
Only education, internship, and publication records already present in the live site or GitHub repositories are shown. Dates that could not be verified are omitted rather than invented.
- education
M.Tech, Applied AI and Machine Learning
VNIT Nagpur
2024 — May 2026
CGPA 6.53/10
- education
B.E., Computer Science
Dr. D.Y. Patil Institute of Technology, Pune
2019 — 2023
CGPA 7.89/10
- experience
Software Engineer Intern
Salesforce (Remote)
Feb 2022 — May 2022
- Built Python bulk data-processing pipelines
- Integrated Salesforce APIs into backend workflows
- Wrote SQL queries with joins and aggregations for CRM analysis
- Built Power BI dashboards for stakeholders
- research
Android botnet detection research (SVM, 342 static app features)
IJRASET 2023
2023
DOI: 10.22214/ijraset.2023.49506 — as listed on github.com/sumit966/-android-botnet-detection
About
How I like to build
I care about systems that stay understandable after the demo: data that can be versioned, models that can be evaluated, APIs that fail loudly, and interfaces that do not hide complexity behind decoration.
The work on this site is organised by hiring role so a reviewer can see the relevant repositories without wading through every experiment. Each project card is backed by a public GitHub repository inspected for this rebuild.
Interests that show up repeatedly in the code: retrieval-augmented generation, computer vision prototypes, analytics dashboards, and small production-shaped backend services.
Contact
Let’s collaborate
Open to internships, applied research, and product engineering roles across AI, data, and software. Reach out with a problem, a dataset, or a product brief.