Job Description:

Technical Core Responsibilities:

1. Agent Architecture & Orchestration:

- Design & Develop: Build sophisticated AI agents capable of autonomous reasoning, task planning, and execution.

- Multi-Agent Coordination: Implement complex orchestration patterns using frameworks like LangGraph, AutoGen, or CrewAI to manage hand-offs and collaboration between specialized agents.

- Tool Integration: Develop seamless "tool-use" (function calling) capabilities, allowing LLMs to interact with external APIs, databases, and legacy software.

2. Prompt Engineering & Reliability:

- Optimization: Create and refine advanced prompting strategies (Chain-of-Thought, ReAct, Reflexion) to improve agent reliability.

- Structured Outputs: Ensure system stability by enforcing structured outputs (JSON, Pydantic) to bridge the gap between stochastic LLM responses and deterministic code.

3. AI Operations (AIOps):

- CI/CD for ML: Design and operate automated pipelines specifically for AI/ML systems, including versioning of prompts, models, and agent configurations.

- Monitoring & Evaluation: Implement "LLM-as-a-judge" or traditional evaluation frameworks to monitor agent performance and drift in production.

Required Technical Stack:

- Languages: Expert-level Python (AsyncIO, Pydantic, FastAPI).

- Frameworks: Deep experience with LangChain and at least one agentic framework (LangGraph is highly preferred for cyclic workflows).

- LLMs: Hands-on experience with frontier models (OpenAI o1/GPT-4, Claude 3.5 Sonnet, Gemini 1.5 Pro) and their specific tool-calling nuances.

- Infrastructure: Experience with Docker, Cloud providers (Azure/AWS/GCP), and modern CI/CD tools (GitHub Actions, GitLab CI).