Hi, I'm Pratik Rathod
AI/ML Engineer with 2 years of experience building production-grade intelligent systems across Computer Vision, NLP, and Generative AI. Skilled in LLMs, RAG, AI Agents, FastAPI, and AWS.
01 — About
Who I Am
I'm a passionate AI/ML Engineer with 2 years of hands-on experience building and deploying production-grade intelligent systems that solve real business problems.
My expertise spans across Computer Vision, Natural Language Processing, and Generative AI — from designing RAG pipelines and multi-agent systems using LangGraph to building CV solutions with YOLO and AWS Rekognition.
At MobileFirst Applications, I led AI/ML initiatives end-to-end — from requirement gathering and solution architecture to client delivery and team mentoring. I thrive at the intersection of research and production.
02 — Skills
Technical Arsenal
Technologies I use to build intelligent, scalable, and production-ready AI systems.
Generative & Agentic AI
Computer Vision
Machine Learning & Deep Learning
Backend & APIs
Cloud, DevOps & MLOps
Databases & Vector Stores
03 — Projects
What I've Built
14 projects spanning Agentic AI, Computer Vision, Generative AI, Research, and MLOps — all production-deployed or published.
SportzVillage
Computer VisionAI-powered sports analytics platform for a sports academy — monitors student performance with real-time video analysis during PT lectures.
- Person detection, gender segmentation, step counting & activity classification
- Fall detection and emotion analysis modules
- Full pipeline with labelled video output and analytics dashboard
SportaJapan
Computer VisionAutomated player detection and recognition with live image/video upload support using AWS Rekognition for facial identity matching.
- AWS Rekognition for face recognition & S3 for face index storage
- Video-to-frame pipeline for accurate player identification
- Deployed on AWS EC2 for scalable real-time usage
XFA TaxAnalyzer
Generative AIW2 tax form analyzer with RAG-powered chat for investment queries, integrated with Nanonet OCR, Stripe subscriptions, and Supabase.
- Nanonet API for structured W2 data extraction (key-value pairs)
- RAG chat for investment and tax query resolution
- Stripe integration for paid subscription plans
RE Analyser
Generative AIReal estate investment chatbot with live property search, deep city analysis, and PostgreSQL-backed conversation history using Brave Search & Perplexity.
- Live property search with Brave Search + Perplexity.ai integration
- Deep city analysis from live + historical data
- PostgreSQL for persistent conversation and city analysis storage
KCO Narration Generation
Generative AIRAG-based text generation system that extracts mandate & matter data from SQL Server, verifies against guidelines, and auto-generates narrations.
- RAG pipeline for mandate/matter data extraction and verification
- Narration editing with compliance scoring against guidelines
- Deployed on Azure with REST endpoints for existing system integration
KCO Insight
Generative AINatural language to SQL analytics platform — converts user questions into structured queries with business logic rules and dynamic chart generation.
- RAG-based SQL pipeline with schema-aware NL to SQL conversion
- Business logic rules layered on top of extracted data
- Dynamic graph generation from query results
CCBST — AuditorAI
Generative AIRAG-based audit automation — extracts structured data from student documents, validates and compares with system data, generates audit reports.
- Text extraction from uploaded DOCX/PDF into structured format
- Validation and reconciliation against system records
- FastAPI endpoints deployed on AWS EC2 + PostgreSQL
Heer — Portfolio AI Assistant
Agentic AIAgentic voice-enabled AI assistant built for this portfolio. Handles multi-intent conversations — from Q&A to structured hiring workflows with JD skill matching and document generation.
- Intent routing with 3 hiring branches via GPT-4o tool calling
- Voice input (Whisper STT) + Text-to-speech (OpenAI Nova TTS)
- Auto-generates skill match reports, project briefs & requirements docs
- Dual SMTP email delivery for every interaction
Financial Agent
Agentic AIAgentic financial analytics system — ingests data from URLs, PDFs, Excel, and CSV; stores in MySQL + FAISS; answers queries via LangGraph agent.
- Multi-source ingestion: web scraping, DOCX, PDF, Excel, CSV
- Hybrid SQL + FAISS vector retrieval via LangGraph tools
- Streamlit interface for natural language financial conversation
Agentic To-Do
Agentic AINatural language To-Do system powered by a LangGraph agent — automates CRUD operations on a MySQL database via schema-aware tool calling.
- LangGraph workflow graph with MySQL database tools
- Schema-based structured output for accurate CRUD automation
- FastAPI backend exposing natural language task management
ML-Genius
ResearchOpen-source Python package (MIT License) on PyPI that automates ML tasks — preprocessing, regression, classification with hyperparameter tuning.
- Auto preprocessing, regression & classification pipelines
- Hyperparameter tuning with best model selection
- Returns reusable model + preprocessor objects; published on PyPI
SQL-Analyser
ResearchNatural language SQL analysis tool using LangChain SQL Agent — answers database questions and auto-generates charts from query results.
- LangChain SQL Agent for NL query resolution
- Automated graph generation with Matplotlib & Seaborn
- Returns analysis in natural language alongside visualizations
Network Security Phishing Classifier
MLOpsEnd-to-end MLOps project classifying network phishing attacks with full CI/CD pipeline using MLFlow, DVC, DagsHub, Docker, and AWS.
- Hyperparameter tuning across Logistic Regression, Decision Tree, Random Forest, SVC
- MLFlow experiment tracking + DVC/DagsHub for data versioning
- Docker + GitHub Actions CI/CD → AWS ECR → EC2 deployment
Customer Purchase Capacity Prediction
MLOpsCustomer segmentation model using K-Means clustering + PCA to rank purchase capacity, deployed with full CI/CD on AWS via Docker and GitHub Actions.
- K-Means clustering + PCA for customer rank segmentation
- Automated preprocessing and training pipeline
- MongoDB Atlas storage + Docker + GitHub Actions CI/CD → AWS EC2
04 — Experience
My Journey
Work Experience
AI/ML Engineer
MobileFirst Applications
- Led AI/ML initiatives and collaborated with cross-functional teams to design and integrate intelligent systems into web applications.
- Built and deployed production-ready AI solutions across Computer Vision, NLP, and Generative AI using Python and FastAPI.
- Contributed to requirement gathering, solution architecture, project planning, and client discussions while mentoring junior team members.
Senior Sales Executive
ICICI Bank
- Assisted customers with mortgage and home loan applications, ensuring efficient documentation and approval workflows.
- Built strong client relationships and generated business through channel partnerships and referrals.
Education
May 2024
M.Sc. in IT (Business Intelligence & Analytics)
Gujarat University
Sept 2020
Bachelor of Commerce
Gujarat University
05 — Certifications
Continuous Learning
Always upskilling to stay at the frontier of AI and machine learning.
Data Science Master
PhysicsWallah (PW-Skills)
MLOps Bootcamp: Build, Deploy & Automate ML with Data Science Projects
Krish AI / Udemy
Data Analytics and Visualization Job Simulation
Forage (Accenture)
Python For Data Science
Cognitive Class (IBM)
Scientific Computing with Python
FreeCodeCamp
SQL and Relational Databases 101
Cognitive Class (IBM)
06 — Contact
Let's Connect
Whether you have an exciting AI project, a collaboration idea, or just want to say hi — I'd love to hear from you. Or just talk to my AI assistant below!
pratikr1521998@gmail.com
Phone
+91 8140577692
pratik-nareshkumar-rathod
GitHub
Pratikrathod08
Designed & built by Pratik Rathod · 2025