Hi, I'm Harjot Singh Raith, an AI and Data Science undergraduate at NMIMS Navi Mumbai (CGPA: 3.2/4.0) specializing in production-grade AI systems, RAG pipelines, Large Language Models, and intelligent scheduling systems.
Currently working as an AI Backend Engineer (Intern) at Paloma POS, USA πΊπΈ, where I design and implement end-to-end production RAG pipelines with Dify workflows, OpenAI embeddings, and Qdrant vector databases, integrated through a centralized AI Gateway (Golang).
- π¬ Research: 2x Research Intern at IIT Bombay (eYantra Lab) with 3 published/accepted papers
- π Competitions: 8+ hackathon wins across 30+ national & international competitions
- π₯ 2nd Place - VOIS International Hackathon (β¬4,000 prize)
- π Judges' Choice Award - eYIC Finals, IIT Bombay (βΉ50,000 prize)
- ποΈ Top 10 - ROBOFEST GUJARAT 4.0 (βΉ2,50,000 for prototyping)
- π‘ Impact: Built production AI systems serving real-world use cases in ed-tech, maritime intelligence & scheduling
- π Leadership: Head of Robotics Club at NMIMS (mentoring 50+ students, organizing technical workshops)
- π Scale: Processed 100K+ educational interaction logs, built multi-tenant knowledge bases
Dec 2025 - Present
- Designed and implemented end-to-end production-grade RAG pipelines using Dify workflows, integrated with a centralized AI Gateway (Golang)
- Implemented low-latency, scalable semantic search using OpenAI embeddings and Qdrant vector database, managing large-scale vector data and retrieval
- Designed robust multi-tenant knowledge base ingestion and AI data infrastructure
- Built automated web scraping, document chunking, embedding lifecycle management, and secure data isolation systems
- Tech Stack: Dify, Golang, OpenAI API, Qdrant, Python, Docker, AWS
May 2025 - July 2025
- Developed a full-stack AI-driven Thematic Analysis platform for multi-format document processing (PDF, DOCX, audio transcripts) with automated inductive and deductive coding
- Built a FastAPI-PostgreSQL backend integrated with LLM-based agents to generate codes, themes, and structured research reports
- Designed a React/MUI frontend with interactive visualizations, including Sankey diagrams, delivering an end-to-end Qualitative Research tool for accelerated insight extraction
- Impact: Reduced manual coding time by 60%, serving 100+ researchers
- Tech Stack: Python, FastAPI, PostgreSQL, React, MUI, LangChain, OpenAI API
May 2024 - July 2024 | Certificate
- Analyzed large-scale user interaction logs (100K+ entries) using Python, leveraging Google OCR for video log generation and CNNs to detect and filter low-quality videos
- Collected, cleaned, and integrated log data via REST APIs from a Laravel-based platform to measure scaffold availability and user dwell time
- Applied data visualization and behavioral analytics to optimize scaffolding and user engagement using clustering, heatmaps, sequence analysis, and predictive modeling (87% accuracy)
- Published research presented at Annual ACM India Compute Conference (Compute 2024), Springer Nature
- Tech Stack: Python, TensorFlow, OpenCV, Google OCR, Pandas, Matplotlib
Published | Annual ACM India Compute Conference (Compute 2024), CCIS vol. 2400, Springer Nature
- Authors: Suprabha Jadhav, Parth Jain, Harjot Singh Raith, Sridhar Iyer, Kavi Arya
- Video analysis and OCR techniques for educational technology and collaborative learning
π YTubeRAG: Leveraging YouTube Content for Enhanced LLM Training with Optimized Vector Space Retrieval
Presented | NTAI 2025, CRC Press (Taylor & Francis Group), Scopus-indexed
- Authors: Harjot Singh Raith, Kartikeya Mudliyar, Ishika Mohan, Aditya Kasar, Sakshi Indolia
- Novel RAG approach for training LLMs using YouTube video transcripts and optimized retrieval
Accepted for Publication | IEEE ICESIC 2026 Conference Proceedings
- Authors: Harjot Singh Raith, Kartikeya Mudliyar, Toral Shah
- Intelligent scheduling system using constraint optimization algorithms and ML
π΄ LiveWire - Real-time Code Compiler (AWS Global Vibe)
Impact: Real-time code compilation platform for developers with AI assistance
- Built a real-time code-compiler platform supporting 10+ programming languages using Monaco Editor, WebSockets, and Gemini AI
- Implemented collaborative IDE with seamless multi-user synchronization for pair programming
- Integrated AI-powered code suggestions and debugging assistance
- Tech Stack: TypeScript, Vite, Tailwind CSS, Monaco Editor, Socket.IO, Express, Node.js, CodeMirror, Gemini API, Vercel, Railway
π ArgoMindAI - Argo Float Dashboard & MapBot (SIH'25 - Smart India Hackathon)
Impact: AI-powered maritime intelligence system with ML-driven ocean disaster predictions
- Developed an automated pipeline to ingest FTP raw files, store them in AWS S3, and generate a sequential schema for real-time dashboards
- Built a map-based RAG system using Argo oceanographic data with ML-driven ocean disaster predictions (85%+ accuracy)
- Created interactive dashboards with React and Leaflet.js visualizing maritime patterns and threats
- Tech Stack: React, Python (UV), FastAPI, Xarray, Tailwind CSS, Leaflet.js, AWS S3, AlloyDB, Docker, Gemini API
π AlmanacAI - Timetable Generation & Optimization (Capstone Project)
Impact: Automated timetable generation reducing manual scheduling time by 90%
- Created a GA/CSP/backtracking-based optimized timetable generator handling 1000+ scheduling variables
- Built RAG system for quick class and faculty lookup using natural language queries
- Developed full-stack platform with constraint optimization algorithms
- Tech Stack: React, Vite, Tailwind CSS, MUI, Axios, Node.js, Express.js, MongoDB, Gemini AI, Vercel, Render
π BIROS - Bio-Inspired Robotic Snake (Gujarat Robofest'25)
Impact: Autonomous search & rescue robot for confined space exploration
- Created an autonomous robotic snake using CNNs, inverse kinematics, and LiDAR for detection in confined and disaster areas
- Implemented serpentine locomotion with realistic snake-like movement using forward and inverse kinematics
- Integrated LiDAR sensors for real-time environment mapping and obstacle avoidance
- Tech Stack: ESP32, Servo Motor MG995, VL53L0X LiDAR Sensors, OpenCV, Python, TensorFlow, Inverse & Forward Kinematics
π₯ 2nd Place - VOIS International Hackathon 2024 by Vodafone | Certificate
Secured 2nd place out of 8000+ teams from India, Romania & Egypt β’ Awarded β¬4,000
π
Judges' Choice Award - eYIC 2023-24 Finals, IIT Bombay | Certificate
Competed out of 299 teams in the finale β’ Awarded βΉ50,000
ποΈ Top 10 Finalist - ROBOFEST GUJARAT 4.0, Science City Ahmedabad | Certificate
Top 10 teams out of 5000+ teams β’ Awarded βΉ2,50,000 for proof-of-concept and prototyping
π₯ 3rd Place - Smart Innovation Model Challenge (BASIC 4.0), VES | Certificate
Secured 3rd place out of 50+ teams β’ Awarded βΉ15,000
π₯ 3rd Place - Startup Pitch Competition (BASIC 4.0), VES | Certificate
Secured 3rd place out of 80+ teams β’ Awarded βΉ15,000
π
Consolation Prize - 9th Multidisciplinary International Conference, Mumbai | Certificate
Paper on LLMs for Indian vernacular speech recognition
π― Multiple Hackathon Wins | All Certificates
30+ hackathons participated β’ 8 wins + numerous awards β’ 40% win rate
Deep Learning β’ Machine Learning β’ Discrete Mathematics β’ Probability and Statistics β’ Computer Networks β’ Cryptography and Network Security β’ Data Structures & Algorithms β’ Object Oriented Programming (OOP)
Narsee Monjee Institute of Management and Studies (NMIMS), Navi Mumbai
B.Tech in Artificial Intelligence and Data Science
July 2022 β April 2026 | CGPA: 3.2/4.0
Shubham Raje Junior College, Thane
Maharashtra State Board, Science
June 2020 β May 2022 | Grade (12th): 85.83%
New Horizon Scholars School, Thane
Central Board of Secondary Education (CBSE)
May 2015 β April 2020 | Grade (10th): 90.40%
- π Cloud Computing | NPTEL, IIT Kharagpur | Certificate
- π€ Generative AI Language Modeling with Transformers | IBM | Certificate
- π Introduction to Generative AI | Google Cloud | Certificate
- π Applied ChatGPT for Cybersecurity | Infosec | Certificate
- π AI Workflow: Enterprise Model Deployment | IBM | Certificate
- π§ Foundation of AI and Machine Learning | Microsoft | Certificate
- π Python Fundamentals | DataCamp | Certificate
- π SQL Fundamentals | DataCamp | Certificate
July 2024 - July 2025
- Conducted workshops on Git, embedded systems, full-stack architecture, and ML-based RAG pipelines for all engineering students
- Organized "Open Source to Deployment: A Full-Stack Journey" workshop series
- Provided hands-on mentorship in robotics, applied AI, and deployment workflows
- Promoted open-source contribution and collaborative development through technical training sessions
- Taught mathematics and science to underprivileged students
- Focused on foundational concepts and problem-solving skills
- Developed engaging lesson plans to improve student comprehension and academic performance in STEM subjects
July 2024 - July 2025
- Led a team of 50+ members in organizing technical workshops and robotics competitions
- Coordinated inter-college events to promote STEM education and hands-on learning experiences
- Managed club operations, event planning, and technical mentorship programs
Sept 2022 - Jan 2023
- Collaborated with team members to organize cultural and technical events
- Contributed to event planning, logistics, and execution to enhance student engagement and participation
π Currently Open For:
- Full-time opportunities in AI/ML Engineering & Research
- Research collaborations in LLMs, RAG systems, and AI applications
- Challenging AI/ML projects with real-world impact
- Speaking engagements and technical workshops
π§ Contact: harjotraith47@gmail.com
π Phone: +91-9321642540
πΌ LinkedIn: Harjot Singh Raith
π Portfolio: harjot.info
π Resume: Download CV
π Location: Thane, Maharashtra, India
β° Availability: Immediate (Ready to start within 2 weeks)

