Description

You will design, build, and scale async REST and WebSocket APIs using Python 3.11+ and FastAPI, implementing multi-agent workflows with Semantic Kernel. You are responsible for integrating LLM providers behind a provider-agnostic layer, delivering Retrieval-Augmented Generation (RAG) systems, and exposing tool-using agents via OpenAI Assistants.

This role is remote.

Responsibilities

• Design and scale async APIs with Python, FastAPI, and clean vertical-slice architecture.
• Implement multi-agent workflows using Semantic Kernel to route traffic among specialized LLM agents.
• Build RAG systems using vector stores such as Azure AI Search, pgvector, or Chroma.
• Expose tool-using agents via OpenAI Assistants for data analysis and file manipulation tasks.
• Maintain CI/CD pipelines (Bitbucket Jenkins) and instrument services with OpenTelemetry and structlog.

Required Skills

• 5+ years building production APIs in Python; 2+ years with FastAPI or similar async stacks.
• Deep knowledge of async I/O, Pydantic v2, dependency injection, and observability.
• Hands-on experience with Semantic Kernel or comparable agent frameworks.
• Practical RAG implementation skills using Azure AI Search, pgvector, or Chroma.
• Strong Postgres skills, including SQLModel/SQLAlchemy 2 and Alembic migrations.
• Proven integrations with LLM APIs (OpenAI, Gemini) and structured-output design.
• Dependency management via Poetry and end-to-end CI/CD ownership.

Preferred Skills

• Experience with event/message queues (RabbitMQ, Azure Service Bus, Kafka).
• Familiarity with observability stacks (Grafana, LangFuse) for LLM cost governance.

Key Skills

Education

Any Graduate