AI Agent & LLM Engineer

About us Raushang4 Technology builds custom AI agents, full-stack products and headless e-commerce for Indian D2C brands, plus Salesforce, IoT and mobile work. We're an early-stage, founder-led company. Custom AI agents are the core of what we do, and you'll work directly with the founder, Raushan Kumar (Full Stack & Agentic AI Architect). What you'll do • Build production agents for D2C support and operations: tools, guardrails, retrieval and evals. Not a demo chatbot. • Build agents that do real work: answering customers on WhatsApp, tracking orders, recommending products and automating ops. • Design the agent architecture and connect tools and data (Shopify, CRMs, catalogues). • Prove quality with evals before anything goes live, and watch reliability and cost as closely as the prompts. Must-have skills • LLM APIs and tool/function calling • RAG or evals on a real project • Python or TypeScript • Clear written architecture (we ask for your GitHub profile when you apply) Nice to have • RAG depth: chunking, embeddings, vector DBs (pgvector/Pinecone/Qdrant) • Agent orchestration (LangGraph, LlamaIndex, Vercel AI SDK or custom) • Guardrails, hallucination checks, cost and latency tracking • Integrations with Shopify, WhatsApp, CRMs and helpdesks • Hindi/Hinglish prompt handling • Observability (Langfuse/LangSmith or similar) How we hire (about 15 working days from application to offer) • Application: apply here on Unstop. • Screening: we review every application within 2 working days. • Practical task: a short task used for assessment only. Paid-trial terms are agreed in writing. • Technical interview: a working session on the kind of problem the role actually ships. • Founder call: a conversation with Raushan about the work and how we collaborate. • Offer.