Design, develop, and deploy advanced Agentic AI systems that autonomously plan, reason, orchestrate tools, and execute multi‑step tasks. Provide deep hands‑on expertise across LLMs, agent frameworks, orchestration layers, retrieval systems, and tool integrations. Lead the development of production‑grade agentic applications while collaborating with architects, ML engineers, and product teams to deliver scalable, reliable, safe, and cost‑efficient autonomous AI systems.
Knowledge & Skills
1.
Strong expertise with LLM ecosystems including OpenAI, Azure OpenAI, Anthropic, Llama, and other transformer-based models; deep familiarity with prompt engineering and agent behavior patterns.
2.
Proficiency in Python and/or TypeScript with hands-on experience in agent frameworks such as LangChain, LangGraph, Semantic Kernel, or custom orchestration engines.
3.
Knowledge of RAG design patterns, retrieval pipelines, vector databases (Pinecone, Chroma, FAISS, Weaviate), and memory systems for autonomous agents.
4.
Understanding of AI safety, guardrails, hallucination mitigation, security controls, versioning, telemetry, and monitoring for production AI systems.
5.
Ability to lead hands-on development, guide engineering teams, review code, troubleshoot complex agent behaviors, and optimize for cost, latency, and reliability.
Mandatory Experience
1.
4–8 years of software development experience with at least 2 years hands-on building LLM/AI applications, autonomous agents, or orchestration-based systems.
2.
Demonstrated experience developing production-grade agentic workflows involving planning, reasoning, multi-step execution, and tool orchestration.
3.
Hands-on expertise integrating LLMs with APIs, structured data sources, vector stores, and cloud-native services across AWS/Azure/GCP.
4.
Experience building robust, modular, scalable AI applications following clean architecture principles, best coding practices, and automated testing.
5.
Strong debugging and optimization skills for LLM-driven pipelines, including prompt refinement, latency optimization, caching, and observability.
Preferred
1.
Experience with multi-agent systems, goal-oriented planning (e.g., ReAct, LATS, OpenAI Swarm-style patterns), or autonomous workflow frameworks.