ABOUT ME PROJECTS EXPERIENCE EDUCATION SKILLS CERTIFICATIONS HOBBIES CONTACT
Your browser does not support the video tag.

HI, I AM SAHIL MOHAPATRA

AI Engineer | LLM · RAG · Agentic AI | Ex-ODeX Global (Dubai) | BITS Pilani CS ’26

ABOUT ME

Sahil Mohapatra

HEY, I AM SAHIL MOHAPATRA,
A COMPUTER SCIENCE UNDERGRADUATE AT BITS PILANI, DUBAI CAMPUS. HERE, YOU CAN EXPLORE THE PROJECTS I HAVE WORKED ON.

PROJECTS

GitFlix

GitFlix

Turns any public GitHub repo into a cinematic documentary with animated scenes, commit analytics, a generated narrative and a full soundtrack.

GitHub Website Demo
Scuba-Doo

Scuba-Doo

Dive safety prediction using ML, weather, marine and live APIs.

GitHub Website Demo
SpendWeiss

SpendWeiss

Multi-agent LangGraph system for card and spend optimisation, streaming live decision traces to an interactive React frontend.

GitHub Website Demo
Social Media Content Analysis

Social Media Content Analysis

Developing a multi-agent AI system that analyses uploaded videos to detect AI-generated or manipulated content and estimate potential virality.

GitHub Website
FootballNotch

FootballNotch

Native macOS menu-bar app turning the camera notch into a live football score overlay.

GitHub Demo
Healthcare AI Assistant using RAG LLMs

Healthcare AI Assistant using RAG LLMs

RAG-based healthcare assistant with agentic appointment routing and query rewriting.

GitHub Website
LLM Fine Tuning

LLM Fine-Tuning with LoRA

LoRA fine-tuning of Falcon-RW-1B using PEFT and TRL.

GitHub Demo
Predictive Analysis

Predictive Analysis of Social Media

ML-driven dashboard for virality prediction, influencer identification, and misinformation spread analysis.

GitHub Website Demo
Music Platform

Music Streaming Platform UI

Responsive Spotify style frontend UI.

GitHub Website

EXPERIENCE

BACKEND AI/ML INTERN

ODEX GLOBAL, DUBAI

AUGUST 2025 – JANUARY 2026

  • Built a production-grade AI document intelligence pipeline for shipping documents using OpenCV, Tesseract OCR and Groq LLMs, converting multi-page PDFs into schema-validated JSON with 97% field-level accuracy.
  • Designed modular Python and FastAPI services for bounding-box detection and context-aware field extraction, integrating them into a scalable production pipeline and significantly reducing manual verification time and operational turnaround.
  • Engineered a multi-stage confidence-scoring framework spanning OCR, classification and extraction layers, aggregating field, document-type and per-field confidence into a single trust score with automatic review flagging for low-confidence fields.
  • Refined LLM prompts to resolve complex multi-BL document hierarchies (Master/House BL logic) and to filter non-data pages, and led a modular codebase restructure (API, services, prompts, messaging and tests) that improved extraction reliability and feature delivery speed.
  • Explored LoRA fine-tuning of Meta-Llama-3-8B and gpt-oss-20b on domain-curated shipping datasets to strengthen layout awareness and field consistency, and prototyped Retrieval Augmented Generation pipelines involving document chunking, embedding generation and vector storage to evaluate contextual grounding for LLM-based parsing.
  • Certificate | Letter of Appreciation

FULL-STACK WEB DEVELOPMENT INTERN

SKOLAR EDTECH, BANGALORE (REMOTE)

JUNE 2024 – AUGUST 2024

  • Developed reusable and responsive UI components using HTML and CSS, standardising layouts and styling to reduce redesign effort and improve iteration speed by over 20%.
  • Built and showcased two complete frontend projects, a Photography Portfolio website and an Online Music Streaming UI , delivering production-ready interfaces that improved client demos and strengthened the company’s project portfolio.
  • Certificate

EDUCATION

BITS PILANI, DUBAI CAMPUS

B.Tech in Computer Science

Batch 2026

DAV PUBLIC SCHOOL, NAVI MUMBAI

Secondary and Higher Secondary

Batch 2022

SKILLS

Languages: Python, Java, SQL, JavaScript, HTML, CSS

Frameworks/Tools: Git, GitHub, uv, Docker, FastAPI, Streamlit, LangChain, LangGraph, Hugging Face, Ollama, PyTorch, TensorFlow, Scikit-Learn, LightGBM, Kafka, MediaPipe, ReAct Agents, MCP Servers

Cloud, Deployment & Databases: AWS S3, Render, Netlify, Supabase, SQL Databases, Vector Databases, FAISS, ChromaDB, REST APIs, JSON Schema Validation, Server-Sent Events (SSE)

AI/ML Areas: Machine Learning, Deep Learning, LLM Fine-Tuning, RAG, Agentic AI, OCR Pipelines, CNN

Core Concepts: Data Structures and Algorithms, RDBMS Concepts, API Design

CERTIFICATIONS

  • 1st Place – ACM Hackathon 2022
  • Complete Generative AI Course With LangChain and Huggingface (Udemy)
  • Complete Agentic AI Bootcamp With LangGraph and LangChain (Udemy)
  • Artificial Intelligence A to Z 2025 (Udemy)
  • Top 5 – ACMW Hackathon 2025

HOBBIES

SCUBA DIVING

Open Water Diver certified, exploring life beneath the surface.

PHOTOGRAPHY AND VIDEOGRAPHY

Capturing stories through unique perspectives.

PIANO

Exploring melodies and enhancing focus.

FITNESS

Regular workouts for physical and mental strength.

TECH ENTHUSIAST

I love doing creative things, you might have seen it in the Projects section, and I keep doing so, follow my GitHub to know!

TRAVEL

Discovering cultures and broadening perspectives.

BUILDING WITH LEGO

Piecing together sets for fun and unwinding, one brick at a time.

CONTACT

LinkedIn Email GitHub X