Generative AI Engineer

Job title: GenAI Engineer — Orchestration, Capability Discovery & Reusable Skills Location: Bay Area preferred (hybrid, 1–2 days/week in office). Remote considered with required overlap of 4–6 hours/day with Pacific Time. About the role We are building an enterprise GenAI application layer that understands user intent, captures screen and workflow context, and routes well-structured requests to existing agents and data services. You will design and implement agent orchestration, routing, and reusable skills; manage tool and capability catalogs; and transform agent outputs into grounded narratives, tables, charts and consistent response payloads. The structured-data query capability exists — this role builds the application logic, integration contracts, and delivery pipeline around it. Responsibilities Design, implement, and maintain agent orchestration, tool-calling, workflow routing, and multi-agent chaining for production systems.Create and enforce structured input/output contracts (JSON schemas) and API interfaces between the application layer and downstream agents/services.Combine user intent, conversation history, screen and workflow context, and filters to produce robust downstream agent requests.Implement capability discovery, reusable skill creation, and registries or catalogs for agents and tools.Transform agent responses into grounded, user-facing outputs: narratives, tables, charts, citations, and consistent response payloads.Build, test, and deploy GenAI services on AWS (including Amazon Bedrock) and AgentCore or equivalent orchestration platforms.Deliver production-ready code with automated tests, monitoring, and clear documentation.Collaborate with product, data, and platform teams to evolve APIs, contracts, and observability for GenAI applications.Minimum qualifications Strong backend engineering experience in Python or TypeScript, with production-grade code and testing practices.Direct production experience with agent orchestration, tool calling, workflow routing, and multi-agent chaining.Proven API and schema design skills; experience defining structured JSON input/output contracts.Experience combining user intent, conversation context, screen/workflow metadata, and filters into downstream agent requests.Demonstrated ability to transform agent results into narratives, tables, charts, citations and consistent application responses.Hands-on experience building and deploying GenAI services with AWS and Amazon Bedrock, and familiarity with AgentCore (or similar orchestrators).Ability to work independently and deliver tested, production-ready features with minimal guidance.Preferred qualifications Experience building capability registries, agent and tool catalogs, workflow skills, or discovery services.Familiarity with MCP (Model-Centric Platform) and API-first GenAI application patterns.Experience building analytics or visualization layers on top of semantic-layer or data-query agents.Knowledge of observability for multi-agent systems: tracing, schema validation, SLA metrics, and auditability.Background in security, data governance, and handling of citations/attributions in generative outputs.What we offer Competitive salary and equity.Health benefits, paid time off, and professional development stipend.Opportunity to shape a foundational GenAI application layer used across the enterprise.Collaborative team with product, data, and platform partners.Interview process (high-level) Recruiter screenTechnical phone screen (system design + coding in Python/TypeScript)Deep technical interview: agent orchestration and API/schema design, sample design challengeOnsite/virtual final loop: implementation walkthrough, architecture discussion, and behavioral interviewHow to apply Submit your resume, a short cover note describing a relevant project (agent orchestration, tool-calling, or capability registry), and links to any public repos or design docs. Include your location and typical Pacific Time overlap.