We are seeking an Agentic AI Engineer to design, construct, and deploy autonomous and semi-autonomous AI agents. These agents will be capable of reasoning, planning, and executing multi-step tasks by leveraging tools, APIs, and external systems. You will be at the forefront of applied AI, building agentic workflows with frameworks such as Google's Agent Development Kit (ADK), LangChain/LangGraph, CrewAI, AutoGen, or similar. Your work will involve integrating these with LLMs like Gemini, Claude, GPT, and open-source models to address complex business challenges.

Design and build sophisticated agentic systems that excel at planning, reasoning, tool utilization, and orchestrating multi-step workflows with minimal human oversight. Develop and deploy agents using leading frameworks like Google ADK, LangChain/LangGraph, Semantic Kernel, CrewAI, or AutoGen. Implement robust integrations for tool-calling, function-calling, and Model Context Protocol (MCP) to connect agents seamlessly with internal APIs, databases, and third-party services. Architect advanced strategies for memory management, state handling, and context management crucial for long-running or multi-turn agent interactions. Construct retrieval-augmented generation (RAG) pipelines and integrate vector databases such as Pinecone, Weaviate, FAISS, or Chroma as required. Evaluate, fine-tune, and select appropriate foundation models, optimizing for cost, latency, and accuracy in specific agentic use cases. Establish comprehensive guardrails, evaluation frameworks, and observability (tracing, logging, eval harnesses) to guarantee agent reliability, safety, and correctness. Collaborate closely with product managers, designers, and domain experts to translate intricate business workflows into effective agent capabilities and task decompositions. Produce production-grade Python (and/or TypeScript) code, adhering to strong testing, CI/CD, and deployment practices for agent services. Continuously stay abreast of the rapidly evolving agentic AI landscape, proactively introducing new techniques, frameworks, and best practices. Document agent architectures, prompt strategies, and decision logic to ensure maintainability and facilitate team knowledge sharing.

A minimum of 3 years of software engineering experience is essential, including at least 1 year of direct experience with LLMs or generative AI systems in a production environment. Proven hands-on experience with at least one agent framework, such as Google ADK, LangChain/LangGraph, CrewAI, AutoGen, or Semantic Kernel, is required. Proficiency in Python is critical, with an emphasis on writing production-quality code. Experience with asynchronous programming is a beneficial asset. A practical understanding of prompt engineering, tool/function calling, and multi-agent orchestration patterns is necessary. Familiarity with integrating LLM APIs (OpenAI, Anthropic, Google Gemini, etc.), including managing rate limits, token economics, and context constraints, is expected. Experience with vector databases and RAG architecture is important. A solid understanding of evaluation methodologies for non-deterministic AI systems, including offline evaluations, human-in-the-loop reviews, and A/B testing, is required. Experience deploying services on cloud platforms like GCP, AWS, or Azure, utilizing containers (Docker/Kubernetes), is a must. A strong grasp of API design, microservices, and event-driven architectures is also necessary.