Staff AI Engineer

Established in 2015, Create Music Group is a leading music and entertainment company. The company operates as a record label, distribution company, and entertainment network which generates over 15 billion music streams each month on DSP’s. Named #2 on the Inc 5000 Fastest Growth Companies in America in 2020, the company has grown exponentially by leveraging its owned IP with its media and technology platform. The company works with superstar artists, major and independent record labels, and global media brands. It operates a number of companies including Label Engine, one of the largest independent music distribution platforms in the world, with over 75,000 artists and 5,000 label clients; and Flighthouse, a digital entertainment brand focused on Gen Z, which has more than 300 million followers across social media. Create Music Group is based in Hollywood, CA and has 400 employees worldwide. Job Summary This Staff AI Engineer role sits at the intersection of AI systems architecture and full-stack product engineering within CMG's AI & ML Engineering org. The person in this seat is the technical anchor for CreateOS's agentic AI layer — owning the design, deployment, and ongoing reliability of the LLM-powered features and agent infrastructure that run across the platform. Day-to-day, this role spans three areas: building and scaling the agent platform (orchestration, RAG pipelines, memory, routing, guardrails); shipping production AI features end-to-end from data model to UI; and serving as the primary technical voice with product, M&A, A&R, and Marketing stakeholders. Unlike a research or prototype role, the bar here is production — systems serving real users at scale, with the observability and reliability expectations of revenue-critical software. Beyond individual contribution, this engineer helps set the architectural patterns and development standards the broader AI team builds against — mentoring the ML Engineer, shaping tooling and primitives, and representing AI Engineering in cross-functional and executive forums. It's a high-autonomy seat for someone who's comfortable deciding what to build and why, not just how. The right candidate brings 6+ years of software engineering experience, at least 3 of which are hands-on with production agentic AI or LLM systems, along with deep fluency in RAG architectures, eval frameworks, and modern AI-native development practices. Responsibilities Agent Platform & LLM Systems Architect AI agents and the orchestration, tool-use, memory, and routing patterns they share — building toward a cohesive agent platform for CreateOSDesign RAG pipelines, retrieval architectures, and semantic search grounded in CreateOS structured data (contracts, royalty statements, catalog metadata, deal terms)Define guardrails, evaluation, observability, and human-in-the-loop standards so agentic systems ship safely and stay measurableDrive model, prompt, and tool-use choices — including cost and latency tradeoffs at production scaleIntegrate frontier LLMs (OpenAI, Anthropic) and selected open-source models into user-facing features across CreateOS modules Stakeholder Partnership & Requirements Partner with product, M&A, A&R, and Marketing stakeholders to turn ambiguous business needs into well-scoped AI featuresRun discovery conversations with internal users to validate the problem before committing engineering investmentTranslate non-technical requirements into clear technical specs; surface tradeoffs (cost, latency, accuracy, scope) early and explicitlyNegotiate scope, timelines, and acceptance criteria with cross-functional partnersDemo work-in-progress regularly; iterate based on real user feedback rather than assumed requirementsDocument decisions and rationale so stakeholders stay aligned across long-running initiatives Production Engineering Ship full-stack AI-native features end-to-end — chat interfaces, copilot tools, workflow automation surfaces — from data model to UIDeploy and maintain production AI services with the reliability, observability, and performance expectations of a revenue-critical systemMaintain CI/CD, testing standards, and code quality for the AI application layerPartner with Data Engineering to consume internal pipelines (dbt, Airflow), third-party feeds (DSPs, distributors), and event-driven flows into product surfaces Technical Direction & Influence Drive architecture decisions for the AI application layer in collaboration with the VP of AI & ML EngineeringHelp shape the patterns, primitives, and tooling the team builds against — so future hires ramp fast and ship safelyMentor the ML Engineer and future AI hires; raise the bar on code review, design docs, and eval rigorRepresent AI engineering in cross-functional forums with product, data, and executive stakeholdersCommunicate tradeoffs clearly across engineering, product, and business audiences Qualifications 6+ years of software engineering experience with a track record of shipping production systems3+ years hands-on building production agentic AI or LLM-powered systems — not prototypes, not demos, systems serving real usersDemonstrated technical ownership of a non-trivial system or platform — you've been the person responsible for the architecture, not just the implementerDeep proficiency in a backend language (Python or Node.js) and working fluency in a modern frontend framework (React, Next.js)Strong experience with RAG systems, vector databases, and embedding-based retrieval — including design tradeoffs (chunking, hybrid search, reranking, freshness)Experience designing eval frameworks for LLM systems — output quality, hallucination detection, regression testing, offline + online evalExperience designing and documenting RESTful APIsProficiency in relational databases (PostgreSQL); comfortable writing and optimizing SQLSolid understanding of containerization (Docker), Kubernetes, and CI/CDProficiency with AI-native development tools (Cursor, Claude Code)Ability to operate with high autonomy and ambiguity — you decide what to build and why, not just how Pay Scale $190,000 - 210,000 USD / yearThe final compensation within this range will be determined based on the candidate’s experience, skills, and overall fit for the role. Details Must be able to attend in person in the Hollywood office 2-3 times per weekInterviews will be a combination of virtual and in-person at our Hollywood Office. Fair Chance Policy In accordance with the Los Angeles County Fair Chance Ordinance, we will consider employment for qualified applicants with criminal histories. We evaluate candidates based on their qualifications and the nature of the offense in relation to the job for which they are applying.