Want to know if this job is worth applying to?
Irvine, California, United States
Hybrid
We're looking for a GenAI Solution Architect who can do both: design the architecture of AI-powered systems and get hands-on building them. This isn't a slide-deck role or a pure advisory role. You'll be orchestrating and building agentic AI systems in real enterprise delivery environments - and helping clients understand what they've actually got.
At CompassX, we're delivering AI-enabled solutions for some of the most complex enterprise clients in financial services, pharma, and beyond. We need someone who has moved past the hype and is actively building with the tools - agents, orchestration frameworks, LLM integrations, and the data infrastructure that makes them work in production.
This role sits at the leading edge of our AI practice. You'll shape how we deliver GenAI engagements and help define what good looks like for our clients.
Architect GenAI Solutions: Design end-to-end architectures for AI-powered applications - from LLM selection and prompt engineering through orchestration, retrieval-augmented generation (RAG), and integration with enterprise data environments.
Build and Orchestrate Agents: Go beyond design - build agentic AI workflows using frameworks like LangChain, LangGraph, AutoGen, CrewAI, or similar. You know how to orchestrate multi-agent systems and have shipped something real.
Own the AI Stack: Work across the full GenAI delivery stack: LLM APIs (OpenAI, Anthropic, Azure OpenAI, AWS Bedrock), vector databases, RAG pipelines, evaluation frameworks, and deployment infrastructure.
Bridge to Enterprise Systems: Integrate GenAI solutions with existing enterprise architecture - data platforms, APIs, identity and access controls, and compliance requirements. Make AI work within the constraints of real organizations.
Advise Client Leadership: Translate technical architecture decisions into clear strategic narratives for client leadership. Help clients move from AI experimentation to AI-enabled operations.
Accelerate the Team: Bring patterns, reusable components, and technical standards to the delivery team. Be the person who makes everyone else faster.
Hands-On GenAI Experience: You have built production or near-production GenAI systems - not just demos. You've wrestled with evaluation, hallucination, latency, and cost at the architecture level.
Agentic AI Fluency: Practical experience building agentic workflows and multi-agent systems using frameworks like LangChain, LangGraph, AutoGen, CrewAI, or similar. You understand where these tools shine and where they fall apart.
LLM Platform Breadth: Hands-on experience with major LLM platforms and APIs: OpenAI, Anthropic Claude, Azure OpenAI, AWS Bedrock, or equivalent. You know the trade-offs between them.
Engineering Foundation: Strong Python skills and comfort in cloud-native environments (AWS, Azure, or GCP). You can write the code, not just spec it.
Data Architecture Context: Understanding of how AI systems consume data - vector stores, embeddings, RAG architectures, and the data infrastructure that sits underneath production AI applications.
Consulting & Communication: Experience working in client-facing or consulting delivery contexts. You can navigate organizational complexity and explain AI architecture to both technical teams and senior business stakeholders.
Builder Mentality: You default to building. You use AI tools to accelerate your own work. You're excited to operate at the frontier of what's actually deployable in enterprise environments - not just what's theoretically possible.
