About
Computer Science (AI & ML) undergraduate shipping full-stack AI products across RAG, OCR, real-time platforms, and applied ML/security tooling.
Jyotiraditya Deb
Computer Science (AI & ML) Undergraduate | Full-Stack AI Builder
about.section
About
Directly adapted from resume content.
I am currently pursuing B.E. in Computer Science Engineering (AI & ML) at M S Ramaiah Institute of Technology (2023-2027).
My work spans FastAPI and Express backends, React and Next.js frontends, graph/vector search systems, OCR pipelines, real-time feeds, and LLM API integrations.
I focus on building practical AI products that combine strong engineering foundations with applied machine learning, security tooling, and clear user-facing workflows.
B.E. in Computer Science Engineering (AI & ML), 2023-2027
Built production-style AI products across Graph RAG, OCR, real-time platforms, and fraud detection
Hands-on with OpenAI, Anthropic, Gemini, Groq, FastAPI, Next.js, PostgreSQL, Neo4j, Qdrant, and Docker
experience.section
Project Journey
Resume-listed projects and ongoing technical explorations.
AI · DATA·Graph RAG System
MedGraph AI
Built a production-style Graph RAG system combining vector search, graph traversal, and LLM synthesis for multi-hop clinical question answering.
- Developed a FastAPI backend with Qdrant and Neo4j for hybrid retrieval workflows
- Built a React/Vite frontend with interactive graph visualization using Sigma, Graphology, and React Query
AI · DATA·OCR Document Pipeline
Doc2Excel AI
Built an OCR-to-Excel pipeline that converts scanned PDFs and images into traceable, structured spreadsheets with review and validation.
- Integrated FastAPI, RapidOCR, PostgreSQL, Celery, and Next.js
- Deployed with Docker, PostgreSQL, Redis, and Celery for batch processing
FULL-STACK·Real-Time Full-Stack Platform
CareerSync
Built a placement management platform with student/admin workspaces, JWT auth, and real-time activity feeds.
- Implemented React, Express, MongoDB, JWT, and Server-Sent Events
- Integrated Gemini/Groq AI career coaching with deterministic scoring fallback and rate-limited recommendations
SECURITY · ML·Applied ML/Security Product
SentinelX Fraud Shield
Built an Android fraud-detection app that monitors call state, caller trust, and voice-stress signals to flag scam-call risk in real time.
- Implemented Kotlin Android client with a FastAPI backend and scikit-learn scoring
- Added LLM-generated explanations and WebSocket-based live dashboard updates
FULL-STACK·Product Engineering
Additional AI Products
Built and explored additional systems including LifeOS Campus, SkillMatch AI, and an AI Trading Agent.
- Designed context-aware focus, cognitive-load, and notification-priority engines
- Built AI roadmap generation and ML signal-scoring workflows
education.section
Education
Academic milestones from the resume.
2023 - 2027
M S Ramaiah Institute of Technology
B.E. Computer Science Engineering (AI & ML)
Completed in 2022
Delhi Public School, Digboi
Class XII
Completed in 2020
Vivekananda Kendra Vidyalaya, Laipuli
Class X
skills.section
Skills & Tools
Technical strengths and language capabilities.
Programming4 items
›Python
›TypeScript/JavaScript
›Kotlin
›C++
Core Concepts6 items
›Data Structures and Algorithms
›Computer Networking
›Graph RAG Systems
›OCR Document Pipelines
›Real-Time Platforms
›Applied ML/Security Tooling
Technologies29 items
›FastAPI
›Express.js
›Node.js
›React
›Next.js
›Vite
›Tailwind CSS
›Framer Motion
›PostgreSQL
›MongoDB
›Neo4j
›Qdrant
›Redis
›Supabase
›Docker
›Docker Compose
›OpenAI SDK
›Anthropic SDK
›Gemini SDK
›Groq SDK
›scikit-learn
›PyTorch
›transformers
›sentence-transformers
›Git
›GitHub
›pytest
›Vitest
›Playwright
Languages4 items
›English
›Hindi
›Bengali
›Assamese
achievements.section
Notable Achievements
Certification and learning milestones.
- Special Mention, Samsung PRISM OpenClaw - Clash of the Claws Hackathon (Jun 2026)
- Represented team Pragmatists with 4 members under faculty mentorship
- Research Paper Presentation - Hybrid Solar-Wind Power System for Reliable and Sustainable Energy Generation at NCETECIME-26, SRM Valliammai Engineering College (Apr 2026)
- AWS Academy Graduate - Machine Learning Foundations
- AWS Academy Graduate - ML for Natural Language Processing
- Google DeepMind AI Research Foundations Track
- Google Cloud Career Launchpad
Interests
Graph RAG SystemsOCR AutomationApplied Machine LearningSecurity ToolingFull-Stack AI ProductsReal-Time Platforms
OPEN · TO · OPPORTUNITIES
projects.section
Selected Projects
Only projects present in the resume.
AI · DATA
FEATUREDMedGraph AI
- Production-style Graph RAG system combining vector search, graph traversal, and LLM synthesis.
- Built with FastAPI, Qdrant, Neo4j, React, and LLM APIs for multi-hop clinical question answering.
- Includes interactive graph visualization using Sigma, Graphology, and React Query.
AI · DATA
FEATUREDDoc2Excel AI
- OCR-to-Excel pipeline converting scanned PDFs/images into traceable structured spreadsheets.
- Built with FastAPI, RapidOCR, PostgreSQL, Celery, and Next.js.
- Supports human-in-the-loop review, validation, batch processing, and ZIP-packaged export.
FULL-STACK AI
FEATUREDCareerSync
- Full-stack placement management platform with student/admin workspaces and JWT auth.
- Uses React, Express, MongoDB, and Server-Sent Events for real-time activity feeds.
- Integrates Gemini/Groq career coaching with deterministic scoring fallback.
SECURITY · ML
FEATUREDSentinelX Fraud Shield
- Android fraud-detection app monitoring call state, caller trust, and voice-stress signals.
- Built with Kotlin, Android, FastAPI, and scikit-learn.
- Includes six-signal ML scoring, LLM-generated explanations, and WebSocket dashboard updates.
AI PRODUCT
LifeOS Campus
- Context-aware student focus and attention system.
- Includes cognitive-load and notification-priority engines.
- Built with Python, FastAPI, and React.
AI PRODUCT
SkillMatch AI
- AI roadmap generator for personalized student income plans.
- Built with React, Express, Anthropic SDK, and Supabase.
AI PRODUCT
AI Trading Agent
- Streamlit dashboard for ML signal scoring, backtesting, and portfolio analysis.
- Built with Python, yfinance, scikit-learn, and PyTorch.