Senior AI Developer Agentic AI
Hands on engineering ownership for enterprise grade AI agents and intelligent workflows
Key Responsibilities
Agentic AI Engineering
Design and implement bounded single agent and multi agent workflows using planning, tool calling, state management, human in the loop approvals, memory, and retrieval where each pattern adds measurable value.
Build agent orchestration with LangGraph or an equivalent framework, including resumable execution, checkpointing, idempotency, timeout handling, loop prevention, and deterministic recovery paths.
Develop secure tools and enterprise connectors using APIs, events, and Model Context Protocol (MCP), with typed schemas, authorization checks, validation, retries, and safe failure behavior.
Implement context engineering, structured outputs, prompt and configuration versioning, model selection, and routing patterns that balance quality, latency, reliability, and cost.
Create reusable agent modules, reference implementations, and integration patterns that can be adopted across multiple business domains.
Knowledge, Retrieval, and Memory
Build and optimize retrieval augmented generation (RAG) pipelines using vector, keyword,
hybrid, or graph retrieval; implement metadata filtering, grounding, citations, and relevance controls.
Engineer ingestion and indexing pipelines for structured and unstructured enterprise content, including parsing, chunking, enrichment, embedding, access control, freshness, and deletion handling.
Select and implement appropriate conversational, episodic, and long term memory patterns while enforcing privacy, retention, tenant isolation, and data minimization requirements.
Evaluation, Safety, and Observability
Define acceptance criteria and build offline and online evaluation suites using representative datasets, golden traces, automated graders, human review, and regression gates.
Measure task success, answer groundedness, tool call accuracy, policy compliance, latency, token consumption, cost, and failure patterns; use evidence to drive iterative improvements.
Implement safeguards for prompt injection, sensitive data, unsafe output, unauthorized tool use, excessive autonomy, and model or dependency failure, aligned with enterprise Responsible AI and security standards.
Instrument agent workflows with logs, metrics, traces, model and tool spans, feedback signals, and dashboards to support debugging, auditability, and production operations.
Production Engineering and Technical Leadership
Develop scalable Python services and APIs using FastAPI, Pydantic, asyncio, and sound
distributed systems patterns; integrate with enterprise applications, data services, and event platforms.
Package and deploy services using containers, Kubernetes, CI/CD, infrastructure as code, and cloud native security controls; contribute to performance, capacity, and reliability engineering.
Apply production resilience patterns including rate limiting*** caching, retries, circuit breakers, fallbacks, concurrency controls, dead letter handling, and graceful degradation.
Partner with product, domain, platform, security, architecture, SRE, data, and application teams to clarify workflows, assess trade offs, deliver integrations, and drive adoption.
Lead code and design reviews, improve engineering standards, investigate complex production issues, document reusable patterns, and mentor other engineers without losing hands on ownership.
Required Qualifications
Demonstrated experience delivering LLM or agent based capabilities beyond prototypes, with evidence of quality evaluation, monitoring, security, and operational support.