Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm Subhash Prajapat, an AI/ML Engineer focused on building scalable, production-grade AI systems using RAG architectures, LLMs, and Generative AI. I thrive on turning complex data into reliable AI products and practical solutions. I've delivered end-to-end healthcare SaaS platforms and real-time chatbot systems, leveraging FastAPI, vector databases, and cloud platforms like AWS and GCP. I excel in data pipelines, semantic search, LLM optimization, and microservices-based deployment to drive performance, scalability, and real-world impact.…I'm Subhash Prajapat, an AI/ML Engineer focused on building scalable, production-grade AI systems using RAG architectures, LLMs, and Generative AI. I thrive on turning complex data into reliable AI products and practical solutions. I've delivered end-to-end healthcare SaaS platforms and real-time chatbot systems, leveraging FastAPI, vector databases, and cloud platforms like AWS and GCP. I excel in data pipelines, semantic search, LLM optimization, and microservices-based deployment to drive performance, scalability, and real-world impact.

Subhash Prajapat

AI Engineer, Data Scientist, Full Stack Developer, +2





I'm Subhash Prajapat, an AI/ML Engineer focused on building scalable, production-grade AI systems using RAG architectures, LLMs, and Generative AI. I thrive on turning complex data into reliable AI products and practical solutions. I've delivered end-to-end healthcare SaaS platforms and real-time chatbot systems, leveraging FastAPI, vector databases, and cloud platforms like AWS and GCP. I excel in data pipelines, semantic search, LLM optimization, and microservices-based deployment to drive performance, scalability, and real-world impact.…I'm Subhash Prajapat, an AI/ML Engineer focused on building scalable, production-grade AI systems using RAG architectures, LLMs, and Generative AI. I thrive on turning complex data into reliable AI products and practical solutions. I've delivered end-to-end healthcare SaaS platforms and real-time chatbot systems, leveraging FastAPI, vector databases, and cloud platforms like AWS and GCP. I excel in data pipelines, semantic search, LLM optimization, and microservices-based deployment to drive performance, scalability, and real-world impact.

Available to hire

I’m Subhash Prajapat, an AI/ML Engineer focused on building scalable, production-grade AI systems using RAG architectures, LLMs, and Generative AI. I thrive on turning complex data into reliable AI products and practical solutions.

I’ve delivered end-to-end healthcare SaaS platforms and real-time chatbot systems, leveraging FastAPI, vector databases, and cloud platforms like AWS and GCP. I excel in data pipelines, semantic search, LLM optimization, and microservices-based deployment to drive performance, scalability, and real-world impact.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Hindi

Advanced

Work Experience

AI/ML Developer at Steve’s AI Lab, Indore, MP, India

April 1, 2025 - November 1, 2025

Developed a multilingual (English & Portuguese) healthcare support AI agent using LangGraph and LLMs to handle patient inquiries, treatment-related questions, and clinic support requests. Built a scalable multi-agent architecture with Intent Routing, RAG-based Knowledge Retrieval, Lead Capture, Incident Management, and Human Escalation workflows. Implemented production-grade RAG pipeline using Pinecone, document ingestion, and contextual retrieval to provide accurate responses from clinic documents, FAQs, and medical knowledge bases. Designed chat conversation management with persistent chat history, memory, user-specific sessions, and secure RBAC. Integrated Redis caching, asynchronous processing, token/cost tracking, and observability pipelines to improve performance, scalability, and operational monitoring. Implemented AI evaluation and monitoring using DeepEval and RAGAS to measure response quality, faithfulness, and retrieval accuracy while applying healthcare data protection, RBA

Junior AI/ML Developer at Nexwen, Bhopal, MP, India

January 1, 2024 - March 1, 2025

Developed an end-to-end machine learning pipeline for vehicle insurance claim prediction. Built a Random Forest model achieving 88% prediction accuracy. Engineered data pipelines using MongoDB and performed exploratory data analysis on large datasets. Developed FastAPI-based prediction APIs and deployed solutions on AWS (EC2, S3). Automated CI/CD workflows using Docker and GitHub Actions.

Content Moderator at TaskUs, Indore, MP, India

April 1, 2022 - October 1, 2023

Reviewed and moderated user-generated content across web and social media platforms to ensure compliance with platform policies and community guidelines. Evaluated text, images, and user interactions (comments, discussions, reviews) to identify and remove inappropriate, harmful, or non-compliant content. Enforced content standards by flagging violations related to harassment, illegal activity, and sensitive material. Maintained high accuracy and consistency in content review while adhering to strict quality and turnaround time targets. Collaborated with internal teams to improve moderation workflows and enhance platform safety.

Lead AI/ML Developer at Shvet Varni Technologies, Indore, MP, India

January 1, 2016 - Present

Developed an end-to-end Healthcare Research RAG SaaS platform using FastAPI, React, and OpenAI models. Engineered scalable document ingestion pipelines for PubMed, Google Scholar, and Semantic Scholar research papers. Architected a vector retrieval system using PostgreSQL, pgvector, and OpenAI embeddings for semantic search. Built Chat-with-PDF functionality using GPT-4o-mini and retrieval-augmented generation (RAG). Designed multimodal pipelines to extract tables and figures from PDFs using Vision LLMs. Implemented JWT authentication, REST APIs, and cloud-based processing workflows for multi-tenant deployment.