3 months ago
Base Salary
$170k - $230k/yr
Responsibilities
- Build agentic workflows over enterprise and government data with model access controls, tool-call rules, and human review or approval points.
- Design context and grounding systems that respect permissions and performance constraints.
- Work across backend services, APIs, asynchronous workers, data pipelines, internal tools, and product-facing surfaces.
- Build evaluations and feedback loops for model behavior and workflow outcomes.
- Own tracing and runtime visibility across models, context, tool calls, and generated outputs.
- Debug production failures using context, traces, tool responses, user review, and logs while improving quality, latency, cost, and security.
- Create reusable primitives for context assembly, grounding, tool use, and reviewable outputs.
- Take ownership of important product and platform surfaces, write maintainable code, and create clear abstractions.
- Use AI coding tools while applying production engineering standards to generated code.
Requirements
- Track record of shipping production software and at least one AI product or workflow used by real users.
- Strong software fundamentals and fluency in Python and/or TypeScript.
- Practical understanding of how models are trained, evaluated, served, and deployed.
- Experience defining access boundaries, measurable quality, failure handling, and debuggable interfaces for AI systems.
- Experience with the stated stack of React/TypeScript, Python services, gRPC, AWS, Kubernetes, Terraform, Snowflake, and Docker is useful, though an exact stack match is not required.
- Strong candidates may have shipped production LLM or agentic workflows, built evaluation or feedback loops, debugged production agent failures, worked with enterprise data security boundaries, or handled high-consequence domains.
Benefits
- Hybrid work arrangement in Seattle, Washington
- Company-paid health coverage, including dependents
- Meaningful equity ownership for early engineers, with flexibility for exceptional scope and impact
- Visa sponsorship is not available
Tech Stack
Categories
About Sobek AI
Sobek AI is the secure infrastructure that lets AI agents work across organizational, national, and IP boundaries. We power co-innovation networks in life science R&D and global health, partnering with research institutes, governments, and pharmaceutical companies on problems no single party can solve alone. We're currently executing on multi-million-dollar contracts spanning international public health coordination and RNA vaccine R&D.