Careers

At Lithos we hunt milliseconds through the whole stack, from GPU kernels to schedulers to serving. If you think our current speed is not fast enough, let's talk.

We are hiring across machine learning, inference optimization, and core systems, along with non-engineering roles across go-to-market and operations. If you do not see the right role but think you belong here, reach out at jobs@lithosai.com

Benefits: Equity, generous health, dental, and vision benefits, plus 401(k) company match.

Visa sponsorship: We sponsor visas on a case-by-case basis.

Engineering

Inference Optimization Engineer

San Francisco, CA or Pittsburgh, PA

Make frontier model inference faster end-to-end. You will work across GPU kernels, compilers, runtimes, batching, scheduling, and serving to turn systems research into production performance.

BS, MS, or PhD in Computer Science or related field.

Apply →

Machine Learning Engineer

San Francisco, CA or Pittsburgh, PA

Make the models themselves faster to serve. You will work on speculative decoding, attention and KV cache optimizations, and model-level techniques that raise tokens per second while preserving full model quality.

BS, MS, or PhD in Computer Science or related field.

Apply →

Infrastructure

Generalist Systems Engineer

San Francisco, CA or Pittsburgh, PA

Build the distributed systems behind fast, reliable model serving. You will work across core infrastructure, resource management, fault tolerance, observability, and APIs as a generalist on the systems that run Lithos at scale.

BS, MS, or PhD in Computer Science or related field.

Apply →

GTM

Developer Relations Engineer

Remote or San Francisco, CA or Pittsburgh, PA

Help developers build faster agents with Lithos. You will create technical content and examples, work directly with users, represent their needs to the engineering team, and grow a community around high performance model inference.

BS, MS, or PhD in Computer Science or related field.

Apply →

Founding Product Marketer

San Francisco Bay Area, hybrid

Be our first marketer. You will own positioning and messaging, create the technical content that earns developer trust, run launches and events, and put our founders in front of the AI developers who care about speed.

Full job description

About Lithos

Lithos builds the fastest agentic inference on the planet: frontier open models at full precision on standard GPUs. Our engine holds top spots on Artificial Analysis for output speed, time to first token, and end-to-end latency on models like DeepSeek and Kimi. We serve inference to growing startups as well as Fortune 500 companies through a public API or on their own hardware.

Lithos was founded by Carnegie Mellon computer science professors who spent years turning systems research on inference and GPU virtualization into a product. We are a small team, and this is one of our first ten hires.

Why this role matters

We have a product that is measurably faster than the alternatives and customers who care about speed. What we don't have yet is a clear story, a known audience, and a steady presence in the places AI developers pay attention.

You will be our first marketer. This is a hands-on role: you will set the strategy and do the work yourself, from positioning to the first blog post. We need someone who can figure out who our buyers are, decide what we should say to them, and get our founders in front of the right people. What you build becomes the foundation for the rest of our marketing.

What you'll do

  • Own positioning and messaging. Define how Lithos is described to AI developers and engineering leaders building agents and customer-facing AI features where speed matters. Turn benchmark results and customer wins into a story people remember.
  • Find our audience. Work out who our best buyers are, where they spend their time, and what makes them try something new. Use what you learn to focus everything else.
  • Create the content that earns developer trust. Technical blog posts, benchmark write-ups, launch announcements, docs pages, case studies, and customer videos, written or shaped by you.
  • Raise the profile of Lithos and its founders. Get the founders into the right conferences, podcasts, panels, and communities, and help them show up prepared, with talks and posts that land.
  • Run launches and events. Plan and ship product launches, and work with partners and investors on developer meetups and executive dinners.
  • Arm the people who sell. Build the pitch, website copy, competitive positioning, and customer-facing materials the founders and our founding account executive use in every deal.
  • Measure what works. Track which channels and messages bring in real usage and pipeline, and put more into what works.

What we're looking for

  • 4+ years in product marketing for a technical product sold to developers or engineering teams, ideally in AI infrastructure, cloud, developer platforms, or data infrastructure.
  • Real technical depth. You can explain latency, throughput and cost per token, read our docs, run a quick benchmark yourself, and hold your own with engineers.
  • Experience at an early-stage startup, where you built the marketing function or a big part of it instead of inheriting one.
  • Your own point of view. You set direction rather than waiting for instructions, and you are comfortable being the only marketer for a while.
  • Excellent writing. You can make a technical idea clear, specific, and interesting in a few sentences.
  • You want to stay hands-on and build, not manage a team right away.
  • Nice to have: you have marketed inference, GPU cloud, model APIs or agent tooling, launched an open-source or developer-first product, or have an engineering background.

How to apply

Email jobs@lithosai.com with your resume, your LinkedIn URL, and a few sentences on a technical product you positioned: who it was for, the message you landed on, and how you knew it worked.

Apply →

Operations

Head of Compute and Capacity

San Francisco Bay Area, hybrid

Own the compute behind Lithos. You will plan capacity ahead of demand, source and negotiate GPU, network, and data-center deals across vendors, and make cost per unit of performance the basis for every capacity decision.

Full job description

About Lithos

Lithos builds the fastest agentic inference on the planet: frontier open models at full precision on standard GPUs. Our engine holds top spots on Artificial Analysis for output speed, time to first token, and end-to-end latency on models like DeepSeek and Kimi. We serve inference to growing startups as well as Fortune 500 companies through a public API or on their own hardware.

Lithos was founded by Carnegie Mellon computer science professors who spent years turning systems research on inference and GPU virtualization into a product. We are a small team, and this is one of our first ten hires.

Why this role matters

For an inference company, capacity is existential. Every promise we make on speed, price, and uptime depends on having the right compute, in the right place, at the right price, before the demand arrives.

You will own that. This is the most important commercial seat at Lithos after the founders, and it reports directly to them.

What you'll do

  • Own the multi-year capacity plan. Translate demand forecasts and the model roadmap into GPU, network, and data-center requirements, and stay ahead of demand rather than chasing it.
  • Run procurement across vendors. Source and negotiate GPU capacity, networking, and colocation across neoclouds, colo providers, OEMs, and hyperscalers. Build a vendor bench deep enough that no single supplier can hold us hostage.
  • Own the commercial terms. Lead contracts end to end: commit profiles, drawdown schedules, pricing, acceptance testing, uptime definitions, service credits, network fabric performance guarantees, and SLA remedies. You will take over live, in-flight deals on day one.
  • Build a rolling price book and vendor scorecard. Track market pricing, delivery reliability, and real-world performance by vendor, region, and GPU generation.
  • Model total cost of ownership (TCO) per unit of performance. Measure cost per token and cost per GPU-hour against what the hardware actually delivers, and make that the basis for every capacity decision.
  • Partner with engineering. Work with our systems and inference teams to make sure the hardware we buy is what our software runs best on, and that acceptance tests prove it.

What we're looking for

  • 7+ years in compute capacity, infrastructure procurement, or data-center commercial roles at a neocloud, colo provider, OEM or channel partner, or a hyperscaler capacity team.
  • You have personally negotiated and closed seven-figure-plus infrastructure contracts, including commit structures, drawdowns, and SLA remedies, and can walk through one in detail.
  • You think and speak in numbers: $/GPU-hour, utilization, power density, delivery lead times, effective cost per unit of performance.
  • Working knowledge of current GPU generations, high-speed networking (InfiniBand/RoCE), and what separates a good cluster from an expensive one.
  • Comfortable as a team of one at first: building the spreadsheets, the processes, and the vendor relationships yourself.
  • Nice to have: experience with capacity for AI inference or training workloads, or on the vendor side selling GPU capacity to AI labs.

How to apply

Email jobs@lithosai.com with your resume, your LinkedIn URL, and a few sentences on the largest compute or data-center contract you have personally negotiated: what you bought, the commit, and the term you are proudest of.

Apply →