Model Engineer - Member of Technical Staff

Job description

Build the AI infrastructure layer of the physical world At Meter, we’re doing something very few teams can: applying frontier AI to reinvent how the internet itself is built, monitored, and managed. We’ve achieved vertical integration of the entire enterprise 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 Fortune 500 companies, schools, factories, and cloud-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 meter.ai https://www.meter.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 vision alongside our executive team. - Real-world systems + frontier AI: This isn’t chatbot optimization. You’ll be building models that power fundamental infrastructure end-to-end, where reliability, accuracy, and latency really matter. - Unmatched data advantage, control over the full-stack: Meter owns the network from physical cables to packet-level telemetry to logs. No vendor or lab has this kind of data or control. - Thousands of H100s at your disposal https://www.meter.com/blog/meter-microsoft: 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, Microsoft, and more. We’re past product-market fit and entering scale. WHAT YOU MIGHT WORK ON - Train end-to-end models, applied to fault prediction, network state modeling, and autonomous repair. - Build multi-modal over structured networking data, performing function-calling to make all decisions on a network . - Evaluate model performance over real-world hardware and virtualized environments. YOU’LL THRIVE HERE IF YOU - Have built, 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.