AI Engineer
Company: PayU Finance India | Experience: 4–7 years | Location: India — Hybrid (Gurugram / Bangalore)

| Type: Full-time, IC`

About the Role

We need an AI Engineer who ships production AI systems — not prototypes. You'll build AI-powered backend services, autonomous agents, and intelligent integrations that directly impact lending decisions and operational e ciency.

You own the full lifecycle — from evaluating a new LLM capability to deploying an agent that handles real tra c with real financial consequences.

Prior hands-on experience building and deploying AI-powered systems in production is a must.

 

 

What You'll Do

Build AI Systems

 

• Deploy AI agents for lending workflows — document verification, bureau analysis, risk assessment, customer communication
• Build MCP servers and tool integrations extending AI across internal systems
• Implement RAG pipelines over lending policies, regulatory docs, and knowledge bases
• Develop function-calling and tool-use patterns for safe LLM interaction with databases and APIs
• Build evaluation and guardrail frameworks to catch hallucinations and enforce compliance

 

Engineer the Backend

 

• Architect scalable services using Java, Spring Boot, and Python for production AI features
• Design schemas using SQL and NoSQL databases optimized for AI workloads
• Build RESTful APIs serving model inference at low latency and high throughput
• Integrate with credit bureaus, banking partners, and payment gateways Raise the Bar

Code reviews for AI patterns and security, optimize token costs and latency, establish AI observability, and drive AI-assisted development (Cursor, Copilot) across the team.

Review and validate AI-generated code for correctness, security, and project standards. Build test suites, guardrails, and eval frameworks for prompts and agent behaviors — with continuous benchmarks, drift detection, and human-in-the-loop review before production.

 

Required Skills

Backend Engineering

 

Area	Expectation
Languages	Java, Python
Frameworks	Spring Boot, Spring Data JPA, Spring Security; FastAPI or Flask
Databases	SQL (schema design, optimization, indexing) and NoSQL (document, key-value, graph)
APIs & Architecture	RESTful APIs, microservices, domain-driven design, event-driven patterns
Cloud	AWS (Bedrock, SageMaker, Lambda), GCP (Vertex AI), or Azure (OpenAI, Cognitive Services)
DevOps	Git, Maven/Gradle, Docker, CI/CD pipelines

AI / ML Engineering

 

Area	Expectation
LLM Integration	Claude, GPT-4, Gemini — API usage, structured output, context management
AI Agents & MCP	Built agents with planning, tool-use, memory; developed MCP servers and tools
RAG & Vectors	Retrieval-augmented generation — chunking, embedding, reranking, vector search
Prompt Engineering	System prompts, few-shot, chain-of-thought, function calling, output parsing
Evaluation	Eval pipelines — accuracy metrics, human-in-the-loop review, regression detection
AI Dev Tools	Daily use of Cursor AI, GitHub Copilot, or equivalent

 

Education