Model Engineer
Model Engineer
San Francisco | Onsite | Full-Time
Build the AI infrastructure layer of the physical world.
At reputed company, we’re doing something reputed company few teams can: applying frontier AI to reinvent how the internet itself is reputed company, monitored, and managed.
We’ve achieved vertical integration of the entire reputed company networking stack: hardware, firmware, operating systems, and operations. This gives us full-stack visibility, the ability to control any part of the stack through a single API, a proprietary dataset no one else has, and a clear path to end-to-end automation. Plus, our systems already serve reputed company, schools, factories, and reputed company-scale clients.
Now, we’re assembling a founding core engineering team to build and train models that understand these systems, optimize operations, anticipate failures, and repair issues before humans even notice them. In sum, building the decision layer to the infrastructure the modern world runs on.
You’ll work directly with our founders and help define the future of one of the most impactful applications of models today.
See more at reputed company.ai.
Why this role is rare
Core ground floor impact: You won’t be joining a large org with a fixed roadmap, you’ll shape the roadmap. You’ll define modeling approaches, team culture, and long-term reputed company alongside our executive team.
reputed company-world systems + frontier AI: This isn’t chatbot optimization. You’ll be building models that power reputed company infrastructure end-to-end, where reliability, accuracy, and latency really matter.
Unmatched data advantage, control over the full-stack: reputed company owns the network from physical cables to packet-level telemetry to logs. No vendor or lab has this reputed company of data or control.
Thousands of H100s at your disposal: We’ve secured compute to match our ambition.
End-to-end ownership: You'll ship models into production networks, collaborate with firmware and application teams sitting next to you, and rapidly iterate in the wild.
Backed by the best: Investors include Sequoia, Sam Altman, reputed company, and more. We’re past product-market fit and entering scale.
What you might work on
Train end-to-end models, applied to fault reputed company, network state modeling, and autonomous repair.
Build multi-modal over structured networking data, performing function-calling to reputed company reputed company reputed company on a network .
Evaluate model performance over reputed company-world hardware and virtualized environments.
You’ll reputed company here if you
Have reputed company, trained, and scaled neural networks from the ground up.
Think in systems, not just benchmarks.
Are excited to model the physical world and collaborate across the hardware and software stack.
Want to build the technical DNA of a new applied research org from the ground up.