AI/ML Engineer – Agentic AI

• Agent Architecture & Orchestration 

• Design and implement agent architectures (single-agent and multi-agent) with robust planning, tool use, and state management. 

• Build orchestration patterns such as supervisor/worker, router-based specialization, and iterative refinement loops. 

• Develop reusable agent frameworks including prompt templates, tool schemas, and policy-based controls. 

• Tooling, Integrations & Automation 

• Create tool interfaces for internal services and data sources (APIs, databases, ticketing, knowledge bases) with strong typing and validation. 

• Implement safe execution patterns (sandboxing where appropriate, permission gating, step limits, and deterministic fallbacks). 

• Integrate agents into user-facing and backend workflows (chat, copilots, automation pipelines). 

• Reliability, Safety & Guardrails 

• Implement guardrails for tool use, data access, and response policies (PII handling, prompt-injection resistance, output constraints). 

• Build monitoring for agent behavior: tool-call success rates, failure modes, loops, latency, cost, and user satisfaction signals. 

• Run incident response and post-mortems for agent failures; improve robustness via systemic fixes and runbooks. 

• Evaluation & Continuous Improvement 

• Design evaluation suites for agent behavior (task success rate, tool correctness, factuality/grounding when retrieval is used). 

• Build regression testing and canary releases to safely ship updates to prompts, tools, and models. 

• Develop feedback loops using user signals, targeted labeling, and automated test generation for recurring failure patterns. 

• Performance & Cost Optimization 

• Optimize agent latency and cost using caching, memoization, selective tool calling, context management, and lightweight models where appropriate. 

• Implement rate limiting, retries, circuit breakers, and queueing strategies to protect downstream dependencies. 

• Collaboration & Documentation 

• Partner with product and engineering teams to translate business workflows into agent designs and measurable success criteria. 

• Document patterns, best practices, and reference implementations for teams adopting agentic systems. 

Required Qualifications 

• Bachelor's degree in Computer Science, Engineering, Data Science, Human-Computer Interaction, or a related field with 5+ years of relevant experience; OR a Master's/PhD with 3+ years of relevant experience. 

• Strong programming skills in Python and experience building LLM-powered applications with tool/function calling. 

• Experience designing APIs/integrations and building secure, maintainable services. 

• Understanding of reliability engineering concepts (observability, incident response, safe rollouts). 

• Experience implementing structured outputs (schemas), validation, and error-handling for production systems. 

• Strong communication and ability to work effectively in cross-functional teams. 

Preferred Qualifications 

• Experience with multi-agent orchestration patterns (supervisor/worker, planner/executor) and stateful workflows. 

• Experience with prompt injection defenses, safety policies, and data governance for enterprise AI. 

• Experience with evaluation frameworks for agentic systems (task benchmarks, simulation, golden tasks, human-in-the-loop review). 

• Experience integrating retrieval (RAG) into agents for grounded reasoning and citations. 

• Experience with workflow engines/queues (e.g.