Agentic Al Engineer(Onsite)(Python coding is must)

Position:Agentic Al Engineer

Location:Minneapolis, MN

Duration:Long term

Job Description:

Python coding along with AI

Programming: Python and TypeScript.

Reasoning Engines: Deep experience with frontier models like GPT-40, Claude 3.5, and Llama 3.x/4 specifically for tool-calling and JSON-mode outputs.

Infrastructure: Experience with Docker, Kubernetes, and Al-specific observability tools like LangSmith

Protocols: Proficiency in Model Context Protocol (MCP) for universal toot connection.

Experience

4+ years in Software/Al Engineering, with at least 1-2 years specifically focused on Al / Autonomous agent workflows

Key Responsibilities

Agentic Orchestration: Design and deploy multi-agent systems using frameworks like LangGraph, CrewAl, or AutoGen where specialized agents collaborate on complex tasks.

Reasoning & Planning: Implement advanced cognitive architectures (eg., ReAct, Plan-and-Execute, Chain-of-Thought) to ensure agents can decompose high-level goals into actionable sub-tasks.

Tool Integration (MCP): Build and maintain secure interfaces using the Model Context Protocol (MCP) to allow agents to interact with internal SQL databases, CRMs, and enterprise APis.

Agentic RAG 2.0: Develop "iterative retrieval" systems where agents autonomously decide if they have enough information or if they need to perform additional searches/queries.

Memory Architecture: Implement long-term and episodic memory systems using Vector Databases (eg, Pinecone, Weaviate) to ensure agents retain context atrossle norajects