Hi, I'm Viral Sherathiya, a Data Scientist and AI/ML Engineer with more than 3 years of experience delivering end-to-end solutions across the full data lifecycle — from ETL pipelines and statistical analysis to production-grade LLM applications and agentic AI systems. I build advanced RAG pipelines, multi-agent LangGraph architectures, and scalable FastAPI microservices on AWS, alongside Tableau and Power BI dashboards that have cut manual reporting by up to 85%.
I specialize in developing production-ready AI applications, LLM-powered platforms, recommendation systems, OCR pipelines, computer vision applications and scalable cloud-based APIs.
I'm Viral Sherathiya, a results-driven Data Scientist and AI/ML Engineer passionate about the full data lifecycle — statistical analysis, predictive modelling, and scalable Generative AI systems.
Over the last 3+ years I've built RAG pipelines, multi-agent LangGraph architectures, recommendation engines, OCR pipelines, and Tableau/Power BI dashboards that turn complex datasets into clear business decisions.
Built AI products using Python, TensorFlow, PyTorch, FastAPI, Flask and Docker.
Developed LLM applications using OpenAI, LangChain, LangGraph, Llama, RAG and Vector Databases.
Created recommendation engines using collaborative filtering, embeddings, metadata, and hybrid ranking algorithms.
Experience with Lambda, EC2, S3, RDS, API Gateway, IAM, Docker, Elastic Beanstalk and deployment pipelines.
Hover a card to see proficiency levels animate.
Over 3 years building AI-powered applications, ML models, recommendation engines, OCR systems and enterprise-grade cloud solutions.
Production-ready Artificial Intelligence, Machine Learning, Generative AI and Cloud solutions.
Architected a modular multi-agent system with specialized LLM agents handling intent analysis, policy Q&A, and insurance-domain conversation flows. Built an advanced RAG pipeline on a FAISS vector database with load balancing, monitoring, and fallback mechanisms. A hybrid local/cloud LLM strategy drove significant reduction in operational cost.
An AI-powered restaurant POS built around six specialized LangGraph agents — Supervisor, Ordering, Modification, Inventory, Payment, and Customer — coordinated with loop-guard logic. Includes real-time inventory management with predictive restocking, deployed as FastAPI microservices via Docker.
Built LSTM and ARIMA forecasting models predicting pharmaceutical sales at 88% accuracy. Delivered interactive Tableau dashboards for stakeholder visualization and supply-chain optimization, with automated retraining pipelines.
Built a transformer-based NLP summarization model with strong ROUGE scores to understand video content, paired with collaborative filtering and content-based algorithms to improve user engagement. Automated CI/CD via GitHub Actions.
End-to-end ensemble ML pipeline using distance, weather, and traffic features to predict delivery times with 92% accuracy, deployed on AWS EC2 with Docker. Power BI dashboards track delivery metrics, driver efficiency, and regional performance.
Jain University · Oct 2021 – Sep 2022
Junagadh Agricultural University · Jul 2018 – Jun 2021
Udemy
Fingertips
Technex, IIT (BHU)
Interested in AI, Machine Learning or Generative AI projects? Let's connect and create intelligent solutions.
+91 81404 08415
Ahmedabad, Gujarat, India
Thanks for reaching out. Viral will get back to you within 24 hours.