Open to opportunities

Rishabh Sharma

@rishab01111

Message

I build production agentic AI systems, RAG applications, and reliable LLM evaluation pipelines.

India

Message

What I'm looking for

I'm looking to build production AI products where agentic workflows, RAG, evaluation, security, and cost control matter. I want hands-on ownership of reliable LLM systems that create measurable outcomes for users.

I've launched a production agentic AI assistant at Kleeto across six HRIS modules, serving 500+ employees and resolving queries the same day rather than in 1–2 days.

I built it with Python, FastAPI, LangGraph, RAG, and 18 MCP tools, pairing evaluation harnesses, human approval gates, and RBAC-scoped retrieval to achieve 85%+ eval-verified task completion with no successful prompt injections in a 50+ case red-team suite.

At Kleeto, I also built a SaaS real-time voice agent for presenting decks and answering questions from organisational documents, plus an automated multimodal onboarding-document verification workflow for two enterprise clients. Previously at Zeonix Global, I launched the company’s first production LLM feature and improved API response time by 40% across three B2B platforms handling 10K requests per day.

I enjoy building measurable, cost-conscious AI products: my Nivara Desk project reached 93.6% correct support-agent decisions across 600 tickets and reduced LLM cost by 26–39% across eight ticket categories without accuracy loss.

Experience

Work history, roles, and key accomplishments

Current

AI Engineer

Current

Kleeto

May 2026 - Present (4 months)

Launched a production agentic AI assistant across 6 HRIS modules for 500+ employees, resolving queries same-day via a Python/FastAPI service with 18 MCP tools, LangGraph, and RAG. Built Kleeto AI Slides, a SaaS real-time voice agent, and automated first-pass verification of onboarding documents for enterprise clients.

Python fastAPI LangGraph MCP RAG LLM Evaluation GuardRails OWASP LLM Top 10 Pipecat WebRTC GPT 4o Docker

Software Developer

Zeonix Global Pvt. Ltd.

Jun 2024 - Apr 2026 (1 year 10 months)

Launched the company's first production LLM feature: a Python/FastAPI microservice with Google Gemini, LangChain and RAG that drafts document-authenticity reports. Improved API response time 40% (1.5 s to 900 ms) on ZeoCRM's highest-traffic endpoints via pagination, GraphQL optimization and PostgreSQL indexing.