Engineer, Backend​/AI Systems

Superset is the AI-native operating system for modern fitness training.

We work with gyms, performance centers, and wellness organizations to deliver personalized training, programming, and health experiences tead of fragmented tools, spreadsheets, and manual workflows, Superset brings everything into a single system of record.

At the core is AI Programming. We use frontier LLMs layered with structured data, rules, and real-world constraints to generate, adapt, and deliver training programs across thousands of members. Coaches stay in control, but the system handles the heavy lifting.

We're design-led, engineering-driven, and building for operators who run real businesses, not just creators or hobbyists.

We're looking for a backend / systems engineer to own core data and infrastructure as we scale.

This is not a typical CRUD full-stack role. You'll be responsible for the foundations that power:

• Large-scale programming across gyms and enterprises
• AI-driven workflows that need to be fast, reliable, and explainable
• Data models that represent real-world training, clients, and outcomes

You'll work directly with the CTO and collaborate closely with product and design. You'll have outsized ownership over architecture decisions and the performance of the system as we scale to thousands of gyms, clubs, and facilities.

• Backend services that power program generation, editing, and delivery
• Background jobs and bulk operations (e.g., pushing programs to hundreds or thousands of users)
• Reliability and performance of the AI pipeline (evals, latency, retries, observability, cost control)
• Data infrastructure and analytics (data lakes/warehouses, pipelines, and modeling for product insights and AI systems)
• Designing systems that balance flexibility with strong constraints and guardrails
• Scaling infrastructure as we grow from pilot projects to enterprise deployments

• Experience building and shipping production systems with real users
• Comfort working on async systems, queues, and data warehouses
• Ability to reason about performance, reliability, and tradeoffs at scale
• Pragmatic approach to architecture. You know when to keep things simple and when to invest
• Interest in AI systems, even if you haven't worked deeply with LLMs yet
• Strong communication and ownership. You can take a problem and drive it end to end

Must have:

• Interest in fitness
• As a member of the team you will be expected to engage regularly in gyms, classes, or structured fitness training
• Immaculate attention to detail and interest in leading projects end-to-end

Nice to have:

• Experience with AI/ML pipelines or LLM integrations
• Experience with high-throughput systems and handling data at scale
• Familiarity with fitness, health, or real-world operational workflows
• Frontend: Type Script, React, React Native (Expo)
• Infrastructure: AWS, Docker
• Data + AI: LLM APIs, structured data pipelines, background workers
• Payments + integrations: Stripe, third-party health and fitness APIs

• You'll own foundational systems that everything else depends on
• You'll work on real AI problems, not demos
• You'll help define how training and wellness programs are delivered at scale

Our founders:

• Peter Coles (CTO, MIT, he/him) has 20 years experience working on and leading engineering teams in NYC including the trail-blazing ML startup, Hunch, which was acquired by eBay, and the innovative wearable company, Ringly. He has a long track record of building and sharing projects on Git Hub and writing about his work, and is a member of the Recurse Center community.
  
  He also built the most popular screens hotting extension, Go Full Page , which is a deceptively simple tool for solving a surprisingly complex problem.
• Taylor Pemberton (CEO, SCAD, he/him) has over 15 years of experience working on brand, design, and product for some of the most inspiring tech companies of today. He ran his own high-end studio that included embedding into product teams in Google to develop a mobile learning platform, Airbnb Experiences, and at Spotify as a product design lead on the creator tools team.
  
  Taylor is NASM certified and a huge gym enthusiast.

We're based in New York City and prioritize candidates with at least ~4 hours of overlap with Eastern Time. Our center of mass is in NYC, but we operate as a distributed team today, with strong representation in Brazil, and are open to candidates in similar time zones.

We believe in having a range of experiences and backgrounds on our team. Diversity makes for a…