Rohit S.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Nov 5, 2026.

I build AI agents, RAG platforms, and LLM-powered systems that stay reliable after deployment—not demos that break in production. 7,000+ Upwork hours • 148 completed projects • Top Rated • 5.0★ across 133 reviews I've delivered production AI solutions for startups and enterprise teams across healthcare, finance, recruitment, manufacturing, SaaS, and e-commerce, focusing on systems clients can confidently operate long after handoff. What I Build AI Agents & Multi-Agent Systems I design and build AI agents that automate business workflows, orchestrate tools, integrate with enterprise systems, and support human-in-the-loop processes using LangGraph, LangChain, Model Context Protocol (MCP), and modern LLM frameworks. Enterprise RAG & Knowledge Systems I develop enterprise knowledge platforms using Hybrid Search, GraphRAG, vector databases, reranking, and RAGAS evaluation to improve retrieval quality, reduce hallucinations, and deliver reliable answers across large document collections. LLM Applications I build AI copilots, document intelligence platforms, workflow automation, research assistants, customer support solutions, and custom LLM-powered applications using OpenAI, Claude, Gemini, and open-source models. Selected Projects Enterprise Multi-Agent GraphRAG Platform Built a production GraphRAG platform connecting enterprise documents, databases, Jira, and Confluence through hybrid retrieval, graph reasoning, and answer verification. RAGAS evaluation improved answer quality by 25%+ while reducing hallucinations across thousands of enterprise documents. HirePilot AI Designed and developed an AI-native recruitment platform handling candidate matching, resume parsing, technical assessments, recruiter verification, interview management, and employer workflows across dedicated candidate, recruiter, employer, and admin portals. The platform reduced manual candidate screening from hours to minutes. MCP Financial Research Agent Built custom MCP servers integrating live market data, financial news, and technical indicators, enabling multi-agent research workflows over real-time financial data and allowing AI agents to perform coordinated market analysis through shared tools. SmartSupportAI Built an intelligent customer-support platform using n8n, Gemini, Pinecone, and cloud knowledge sources that automatically synchronizes manuals, FAQs, and knowledge-base updates, reducing document search time by approximately 75% while minimizing repetitive support requests. Multi-Agent Healthcare Assistant Developed a healthcare assistant supporting department routing, appointment scheduling, prescription explanations, patient-history retrieval, and escalation to human staff. Computer Vision & Document AI Built deep-learning solutions for medical image segmentation, OCR pipelines for architectural and engineering drawings, document understanding, and AI-powered image analysis. Reinforcement Learning Alongside LLM applications, I've built reinforcement learning systems for complex decision-making problems. Projects include a multi-agent dialogue framework where agents learned to negotiate and argue with each other through a custom reinforcement learning layer integrated with an external knowledge base, a self-play chess agent trained against both human players and chess engines, and an RL pipeline for protein structure merging that used an LLM as the reward model to guide learning. This experience gives me a strong foundation in building systems that learn, plan, and optimize autonomously. Technology Stack Programming Languages Python AI & Agent Frameworks LangGraph • LangChain • MCP (Model Context Protocol) • OpenAI • Claude • Gemini • Hugging Face Transformers RAG & Knowledge Systems GraphRAG • Hybrid Search • Pinecone • Weaviate • FAISS • Neo4j • RAGAS Backend & APIs FastAPI • PostgreSQL • Redis • REST APIs Deployment & Infrastructure Docker • AWS • Azure • Google Cloud Platform Automation n8n Why Clients Hire Me I build production systems—not portfolio demos. You work directly with the engineer designing and building your solution. I recommend the simplest solution that solves the business problem, even when it means using less AI. I explain technical trade-offs clearly so you understand why a solution was chosen. Every solution is designed for long-term maintainability, scalability, and practical business use.