Open to opportunities

Sandesh Pathak

@sandeshpathak

Message

Full-stack AI engineer shipping end-to-end LLM-powered product workflows.

Nepal

Message

What I'm looking for

I’m looking to build and ship product workflows powered by language models—owning features end-to-end and focusing on production-grade reliability, strong evals, and responsible AI practices, aligned with teams incorporating LLMs into real work.

I’m a full-stack AI engineer who owns features end-to-end, pairing React/Next.js + TypeScript frontends with Node.js/Python backends to deliver production-grade, LLM-powered systems. I build RAG pipelines, LangChain/LangGraph orchestration, multi-agent frameworks, and MCP servers connected to live CRM and operations data.

I’ve shipped end-to-end automation for a CRM-AI platform serving a 1,500+ employee operation, including billing automation that automated 10% referral-fee reconciliation and reduced reconciliation errors to zero. I also lead responsible AI engineering practices—prompt versioning, eval gates, rollout checklists, and HITL checkpoints—so language-model features reach product reliably and safely.

Experience

Work history, roles, and key accomplishments

AI Engineer

Gauntlet AI

Feb 2026 - Jun 2026 (4 months)

Built LLM-powered systems end-to-end, including a commit planner UI with a Claude layer, a capital-markets RAG training platform, and an event-driven ingest pipeline streamed via SSE. Designed a typed-step multi-agent orchestration engine with LangGraph, along with LLM evaluation harnesses and responsible AI rollout/eval playbooks.

React 18+TypeScript Zustand Server Sent Events (SSE)Spring Boot PostgreSQL JWT fastAPI WebSockets RAG Vector Databases OpenAI API LangGraph Multi agent orchestration Rust Graceful Shutdown CI Quality Drift Detection Prompt Versioning HITL Checkpoints

Software Engineer Lead

All My Sons Moving and Storage

Apr 2022 - Jan 2026 (3 years 9 months)

Led full-stack development of a CRM-AI automation platform, incorporating LLM intent classification and LangChain/LangGraph orchestration to automate customer routing with RAG-based vendor matching. Built production AI integrations including a Django-based MCP server for scoped agent access to CRM data, plus a no-code automation flow designer for operators.

GraphQL MongoDB AWS Lambda SQS Twilio LangChain LangGraph LLM Intent Classification WebSockets MCP Django Scoped Permission Boundaries No Code Automation REST Lang Chain

Software Engineering Intern

All My Sons Moving and Storage

Jun 2021 - Mar 2022 (9 months)

Worked on full-stack development of a CRM-AI automation platform for a 1,500+ employee national operation, including lead intake, intelligent routing, and real-time customer communications.

GraphQL MongoDB AWS Lambda SQS Twilio Full Stack Developer Customer Communications