Job Summary

We are seeking an Agent Engineer to design, build, and optimize reusable AI agents and workflows using Agent Builder, contemporary agent frameworks, and enterprise LLM toolkits. In this role, you will create specialized operational agents for ticket analysis, knowledge mining, SLA tracking, CMDB assessments, and automated runbook generation across transition and BAU environments. You will be responsible for ensuring all agents are accurate, testable, secure, grounded, and fully observable, while leveraging production telemetry, evaluation metrics, and SME feedback to continuously refine agent performance.

Responsibilities

- Agent Architecture & Workflow Design: Translate AMS process requirements into single and multi-agent workflows using Agent Builder, LangChain, LangGraph, AutoGen, or CrewAI.
- Tool & Memory Implementation: Build advanced agent capabilities including tool calling, function execution, persistent memory, reflection patterns, and human-in-the-loop escalation paths.
- Prompt & Context Engineering: Design system prompts, tool instructions, structured output schemas, and dynamic context-assembly logic for optimal response generation.
- Grounding & Guardrails: Minimize hallucinations by strictly grounding agents to approved sources, enforcing safety guardrails, and applying validation checkpoints.
- Evaluation & Optimization: Construct golden datasets, regression tests, and scoring rubrics to continuously benchmark and tune retrieval, accuracy, and usability.
- AMS Use Case Delivery: Develop specialized agents for ticket triage, knowledge mining, SOP/runbook generation, SLA analysis, and incident pattern discovery.
- Production Engineering & Reliability: Package agents for release with version control, unit tests, automated retry logic, error handling, and fallback mechanisms.
- Observability & Continuous Improvement: Implement telemetry, tracing,



and user feedback loops to triage production defects and refine agent performance post-deployment.

Requirements

- Experience : 5+ years in software, automation, or platform engineering, with 2+ years dedicated to hands-on development of GenAI, LLM, RAG, or AI agent applications (AMS/ITSM experience preferred).
- Agent Platform Mastery: Hands-on experience with Agent Builder or enterprise agent development platforms, including workflow design, tool registration, multi-agent orchestration, and deployment packaging.
- Agent Frameworks & Patterns: Proficiency in agentic patterns (planner, executor, critic, router, memory) using frameworks like LangChain, LangGraph, AutoGen, CrewAI, LlamaIndex, or Semantic Kernel.
- LLM & RAG Ecosystem: Strong knowledge of prompt/context engineering, function calling, JSON schema outputs, vector search, embeddings, re-ranking, and model endpoints (OpenAI, Gemini, Claude, Llama).
- Core Software Engineering: Advanced Python skills, REST APIs (e.g., FastAPI), Git, CI/CD, unit testing, and building observable, fault-tolerant distributed workflows.
- Enterprise ITSM Systems: Understanding of AMS/ITSM domains, including ServiceNow data structures (incidents, problems, CMDB, SLAs), ticket lifecycles, and runbook automation.
- Governance & Reliability: Ability to implement deterministic controls, human-in-the-loop approvals, audit logs, and performance evaluation metrics to minimize hallucinations.
- Key Delivery Track Record: Proven experience producing end-to-end agent assets, from design specifications and prompt configurations to production test suites and release guides.

Preferred Technical and Professional Experience

- Master Degree & any relevant certification

Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.

📌 AI Agent Engineer (Pune)
🏢 Kyndryl
📍 Pune