Senior Applied AI Engineer

Alembic is an applied science company building GPU-resident distributed data systems that deliver 10–100x performance for Fortune 500 clients including NVIDIA and Delta. We're Series B ($145M raised), ~70 people, headquartered in San Francisco with a New York office and our SV11 compute facility. Our stack runs on a 256-petaflop NVIDIA DGX cluster with NVL72 GPU infrastructure, combining Spiking Neural Networks, Graph Neural Networks, and causal inference to deliver real-time analytics that were previously impossible.

We're hiring a Senior Software Engineer onto our Applied AI team to build and extend the backend systems that power our platform. This is a hands-on role on a small team where your work ships to production quickly and directly shapes what our largest customers see. You'll work across Python-heavy backend services, data systems, and the infrastructure layer that connects them to our GPU-resident compute.

• →

  5+ years of backend software engineering experience in production environments

• →

  Strong Python fundamentals and experience building and operating backend services

• →

  Demonstrated ability to work across adjacent parts of a stack (data, infrastructure, APIs) rather than staying in a narrow lane

• →

  Track record of shipping in fast-moving, ambiguous environments

• →

  Clear written and verbal communication — you can articulate tradeoffs, explain decisions, and collaborate across functions

• Experience designing and operating distributed systems

• Comfort with performance-sensitive code and systems where latency and throughput matter

• Exposure to data-intensive applications — pipelines, storage systems, or analytical workloads

• GPU or accelerator-adjacent engineering experience

• Background in high-scale or high-performance computing environments

• Experience partnering closely with applied science or research teams

• Familiarity with causal inference or graph-based systems

• Work on systems that are genuinely novel — GPU-resident infrastructure running real-time causal computation at a scale few companies are attempting

• Customers who use the product seriously — NVIDIA, Delta, and others rely on what we build

• Small team, high ownership, short path from idea to production

• Five days onsite in a downtown SF office with a team that cares about the craft