Sr. AI Engineer (Full Stack)

Peloton•New York, NY

3d•$160,600 - $224,950•Hybrid

About The Position

Peloton is seeking a creative, fast-moving AI Engineer to turn ideas into production-grade applications for its global organization. The role architects intelligence and integration layers that connect autonomous agents to Peloton’s enterprise infrastructure. It focuses on shipping useful AI applications quickly, with an emphasis on systems thinking, engineering fundamentals, and product sense. The engineer will build AI features and tools, create memory architectures and vector stores that preserve context while following data retention and deletion policies, and develop pipelines that connect fragmented enterprise systems. The role also works closely with business units and stakeholders to identify valuable use cases, guide adoption, and share feedback about model performance. It includes deploying evaluation pipelines and implementing semantic caching and model routing to improve safety, latency, and inference efficiency.

Requirements

5–7+ years of general software engineering experience.
Strong foundational software engineering skills, including distributed systems, asynchronous patterns, API design, microservices, and databases.
Ability to pivot to the language or framework that gets the job done fastest.
Hands-on experience deploying RAG and agentic workflows into production using LLM APIs, vector databases, and orchestration frameworks.
Track record building and maintaining pipelines across fragmented enterprise systems, with attention to entity resolution, schema design, and data quality.

Responsibilities

Turn ambiguous problems into functional, high-performing AI features and standalone tools.
Build memory architectures and vector stores that maintain context while adhering to data retention and deletion policies.
Build pipelines and entity-resolution logic to connect data across CRM, ERP, CPQ, and marketing systems into a reliable source of truth for agents.
Work with Peloton business units to understand workflows, identify high-value opportunities, and select initial use cases.
Work closely with stakeholder teams, understand their needs, and guide adoption of solutions.
Share field feedback about where models succeed and where they can improve.
Deploy continuous evaluation pipelines to test for hallucinations, prompt injections, data leakage, and performance regressions before production.
Implement semantic caching and model-routing logic to reduce latency and inference costs.