Staff AI Platform Engineer: Agent & Retrieval Infrastructure
Bedrock Ocean Exploration · Remote · Remote ·
COMPENSATION
€•• – ••k
Listed on the employer's job board (Ashby)
- Staff level role
- Fully remote
- Amazon Bedrock focus
- Python or TypeScript
- $160k-$220k base
You will lead the architecture of AI agents and retrieval infrastructure, moving beyond prototype implementations to robust production systems. The work combines software engineering, data engineering, and infrastructure operations to support internal tools and customer-facing ocean data platforms.
Security and governance are central to this role, requiring you to define operational boundaries for autonomous agents. You will build abstraction layers and SDKs that enable other engineers to implement AI features independently while ensuring data isolation and compliance.
What you'll do
- Design Amazon Bedrock integration including agent configuration and backend APIs
- Own end-to-end retrieval pipeline from ingestion to OpenSearch Serverless storage
- Implement security measures like Guardrails, VPC boundaries, and least-privilege IAM
- Build mechanisms to enforce approval boundaries for autonomous agent actions
- Establish LLMOps observability using CloudWatch and tools like Langfuse or Phoenix
- Create automated evaluation frameworks to track model accuracy and release gates
- Provide SDKs and self-service environments for engineering team productivity
- Manage infrastructure as code across all stages for deployment safety
What you'll need
- 8+ years in software and infrastructure engineering with staff-level ownership
- Deep production backend experience in Python, TypeScript, or Go
- Hands-on experience deploying Amazon Bedrock agents and knowledge bases in production
- Experience with containerized services on ECS, EKS, or Lambda and owned CI/CD
- Practical RAG and vector search experience at production scale and cost
- Strong AWS expertise including IAM, VPC, PrivateLink, and infrastructure as code
- Proven track record moving LLM features from prototype to production dependency
- Experience designing developer-facing APIs and multi-tenant services with isolation
Nice to have
- Experience deploying LLM evaluations as release gates
- Involvement in AI red-teaming or security community
- Knowledge of GraphRAG or knowledge graphs
- Experience with geospatial or large-binary dataset retrieval
- Background in handling intermittent data links or edge systems
- Compliance experience such as SOC 2 or government data handling
Stack
- Amazon Bedrock
- Python
- TypeScript
- Go
- Amazon OpenSearch Serverless
- AWS ECS
- AWS EKS
- AWS Lambda
- Terraform
- CloudWatch
- Langfuse
How you'll work
- Fully remote position
- Small team environment requiring versatility
- Focus on shipping via managed infrastructure
What you get
- Base salary $160,000 to $220,000 annually
- Equity compensation included
- Equal opportunity employer