Job Description: Key Responsibilities

1. Rapid Prototyping & Application Development

- Build AI applications, copilots, and agentic workflows end-to-end – UI, APIs, business logic, and model integration.

- Use rapid development tools (Cursor, Claude Code, Replit, Google AI Studio) to compress build cycles and iterate quickly with users and stakeholders.

- Turn loosely-defined requirements into working demos and prototypes within days , then refine based on feedback.

2. Agentic & GenAI Engineering

- Develop with agentic SDKs and frameworks – OpenAI Agents SDK, Anthropic Claude (Agent SDK / API), Google Gemini & ADK, LangChain/LangGraph.

- Implement RAG pipelines, tool/function calling, structured outputs, and prompt engineering with systematic testing and evals.

- Integrate models and agents with enterprise data sources and APIs , handling auth, rate limits, and error paths properly.

3. Engineering Quality & Productionization

- Write clean, testable, well-documented code ; use Git, containers, and CI/CD as standard practice.

- Partner with Forward Deployment Engineers and platform teams to take successful prototypes into production, adding monitoring, guardrails, and cost controls.

- Balance speed and quality pragmatically – knowing when to hack and when to harden.

4. Collaboration & Continuous Learning

- Work closely with architects, data scientists, and designers ; contribute to demos, accelerators, and internal hackathons.

- Stay current with the fast-moving model and tooling landscape , and share learnings across the team.

- Evangelize AI-assisted development practices that raise the whole team’s velocity.

Technical Skills & Tooling (Hands-On)

- Rapid development tools as daily drivers : Cursor, Claude Code, Replit, Google AI Studio, GitHub Copilot – demonstrated ability to ship real software with AI-assisted workflows.

- Agentic SDKs & frameworks : hands-on experience with OpenAI Agents SDK,



Anthropic Claude APIs/Agent SDK, Google Gemini/ADK, and LangChain or LangGraph.

- Strong programming skills in Python and/or TypeScript/JavaScript; comfort building full-stack prototypes (React/Node) and REST APIs.

- LLM application patterns : prompt engineering, function/tool calling, structured outputs, RAG with vector stores (pgvector, Pinecone, FAISS, or similar).

- Testing & observability basics : writing evals, using tracing tools (LangSmith, Langfuse, or similar), and monitoring cost/latency/quality.

- Engineering foundations : Git, Docker, CI/CD, and at least one cloud (AWS/Azure/GCP).

- Good to have : voice/multimodal experience (ElevenLabs, HeyGen), MCP-based tool integration, fine-tuning or open-source LLM experience.

- Responsibilities: Speed of delivery : consistent idea-to-prototype turnaround in days and prototype-to-production in weeks.

- Volume and quality of shipped work : applications, demos, and accelerators that are actually used by stakeholders and internal teams.

- Reliability of what ships : low defect rates, sensible test/eval coverage, and predictable cost/latency behavior.

- Contribution to reuse : components, patterns, and utilities adopted by other engineers.

Team velocity uplift through shared AI-assisted development practices.

- Qualifications: 4–8 years of software engineering experience , with 1–2+ years building GenAI/LLM applications hands-on.

- A portfolio of shipped AI work – products, prototypes, GitHub projects, or demos you can walk us through.

- Bachelor’s degree in Computer Science, Engineering, or related field (or equivalent practical experience).

- Demonstrated fluency with AI-native development tools (Cursor, Claude Code, Replit, AI Studio) in real projects – not just experimentation.

- Solid problem-solving skills and product sense – you care about whether the thing you built actually gets used.

Clear written and verbal communication ; comfortable demoing your work to technical and business audiences.

📌 Senior Manager (India)
🏢 EXL Service
📍 India