Description

You will design and maintain LLM-powered multi-agent workflows for code analysis, remediation, and patch generation.

Responsibilities

• Implement agentic patterns including planning/execution loops, tool orchestration, sandboxing, and failure recovery.
• Architect and optimize RAG pipelines, managing chunking, embeddings, hybrid search, and reranking.
• Build scalable automation systems for technical debt remediation, including language upgrades and vulnerability patching.
• Develop evaluation frameworks for LLMs and agent workflows using offline datasets and statistical testing.
• Contribute to backend systems using Python, distributed systems, microservices, and PostgreSQL.
• Build CI/CD pipelines and observability dashboards to monitor model and network performance.

Required Skills

• 5+ years of experience building and owning production-grade systems.
• Expertise in Python, system design, and software engineering best practices.
• Hands-on experience shipping LLM-powered features, autonomous workflows, or function calling.
• Deep knowledge of multi-agent architectures and RAG systems.
• Proficiency with AWS services including Lambda, ECS/EKS, S3, API Gateway, EC2, and IAM.
• Experience with Infrastructure-as-Code and CI/CD tools like GitHub Actions.
• Knowledge of vector databases such as Pinecone or pgvector.
• Experience with PostgreSQL, DBT, data modeling, and schema evolution.
• Proficiency with observability tools including Datadog, logging, tracing, and metrics.

Preferred Skills

• Experience designing agent infrastructure with sandboxing and tool isolation.
• Background in large-scale platform engineering or developer experience tooling.
• Strong architectural communication skills, including RFC writing and diagramming.

Key Skills

Education

Bachelor's degree