Want to know if this job is worth applying to?
San Francisco, California, United States
Hybrid
Andromeda gives AI companies access to the kind of scaled compute once reserved for hyperscalers. Our platform connects 100+ AI customers to 50+ global providers, with billions of GPU-hours supported, and those numbers are all rapidly growing. We combine enterprise-grade reliability with the speed and economics of an open market, serving teams running everything from large-scale training to production inference.
Nat Friedman (former CEO of GitHub) and Daniel Gross (former head of AI at Apple, YC partner) started Andromeda in 2023 with a single GPU cluster. It filled almost immediately. Three years later, we're a $1.5B company, profitable since day one, with a Series A from Paradigm to scale the platform globally. The global flow of compute is already a multi-trillion dollar market, and our team is building the infrastructure that enables it to continue to scale.
The problem is deceptively hard. Not all compute is equal: interconnect, networking, OEM, firmware, and cluster age all vary across providers, and the differences matter at scale. Our platform benchmarks and validates capacity, takes positions, structures contracts, and operates clusters globally, delivering a consistent product regardless of where it runs. No one else has built this layer, and the AI industry can't scale without it.
We're hiring our first Solutions Architect to own the technical side of our largest customer engagements. You will be the technical face of Andromeda to frontier labs and enterprises evaluating Andromeda to train and serve their models, as well as the advocate for those customers after they sign.
You'll own your accounts end-to-end. You run the technical discovery and the POC that wins the deal, then you're the person who lands the cluster, gets the first real training run through it, and stays the technical owner as the account grows. No handoff to a stranger at signature. The architecture you promised in the evaluation is the architecture you're accountable for in production.
You are the person in the room who can go from a CTO's roadmap to an interconnect topology without losing either audience.
This is greenfield. There's no SA playbook here yet; no demo environment, no evaluation template, no onboarding runbook, no reference architectures. You'll build them, and they'll become how Andromeda works with customers for years. If you've been the person who quietly built all of that at your last company and want to do it on purpose this time, this is that role.
● Partner with Sales to qualify opportunities, lead technical discovery, and scope evaluations against the customer's real success criteria — not a generic checklist.
● Design cluster and workload architectures that map to business outcomes: time-to-first-token, tokens/sec, MFU, cost per training run, reliability targets.
● Own POCs end-to-end — scope, benchmarks, success metrics, timeline, stakeholder alignment.
Land it and grow it
● Take customers from signature to first successful production training run: provisioning, environment setup, validation benchmarks, and the unglamorous debugging in between.
● Serve as the named technical owner for your accounts after launch. You’re responsible for architecture reviews, capacity planning, performance and cost optimization, and technical workload reviews.
● Spot expansion before the customer asks: where they're capacity-constrained, what's coming on their roadmap, and what it will take to serve it.
Build the function
● Create the assets the SA team runs on: demo environments, benchmarking harnesses, reference architectures, onboarding runbooks, evaluation playbooks, and competitive material.
● Be the highest-signal feedback loop into Product, Engineering, and Research — recurring gaps, competitive losses, and what customers actually ask for once they're in production.
● 5+ years in customer-facing technical roles, including 2+ years in pre-sales (Solutions Engineer, Sales Engineer, Solutions Architect, or specialist SA).
● Direct experience selling or supporting GPU compute at a neocloud or GPU provider, or as an AI/HPC specialist at a hyperscaler or NVIDIA.
● Demonstrated ability to take a customer from evaluation into production and stay accountable for the outcome.
● Real fluency in large-scale training and inference: distributed training frameworks, multi-node topologies, InfiniBand/RoCE, storage and checkpointing, and where these break at scale.
● Comfort with Kubernetes and SLURM as scheduling environments customers actually run in.
● Enough Python to build a benchmark, a prototype, or an API integration yourself rather than waiting on engineering.
● A track record of owning technical evaluations in complex, multi-stakeholder deals and changing the outcome.
● Exceptional communication, able to hold a deep conversation with a distributed systems engineer and a CFO.
● Experience with frontier labs or AI-native companies as customers.
● Performance benchmarking, MFU analysis, or total-cost-of-training modeling.
● Having been the first or earliest SE somewhere.
● POCs convert reliably because evaluations are well-scoped, technically credible, and tied to customer ROI.
● Customers reach their first production training run fast, and what they saw in the POC is what they get.
● Accounts grow because you saw the constraint coming before the customer raised it.
● The assets you build let the next five SAs ramp in weeks instead of quarters.
● High-growth environment: Get in early at a company at the center of the AI infrastructure boom
● Competitive compensation: + meaningful equity
● Comprehensive benefits: for you and your dependents, including healthcare, dental, and vision coverage, 401(k), and unlimited PTO
Andromeda Cluster is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees. We do not discriminate on the basis of race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, or disability status.