Engineering Manager; Embedded

DS creates systems that power the next generation of radio spectrum intelligence. We collect radio data from all over the world, train neural networks to decipher it, and run them on the smallest chips we can. We’re solving a new, technically hard problem where nothing from other fields works out of the box, and along the way, we’ve built our own stack from scratch, including entirely new embedding model architectures, custom GPU kernels, and much more.

Joining DS means owning major parts of a fast-growing AI research organization, joining a collaborative, talent-dense team with decades of experience in probabilistic ML, accelerated computing, embedded systems, and signal theory, and growing your career in the areas that interest you. You’ll fit in if you want to come to work for the problem itself and don’t want to choose between technical rigor, business value, and real-world impact.

We work with high ownership and trust.

We are hiring an Engineering Manager (Embedded) to lead the team that builds the software running on our sensors. This is the team accountable for everything that happens on the device: the embedded Linux platform, the interfaces to our hardware, the runtime that executes our models on the smallest chips we can, and the reliability of every unit we have deployed in the field.

This is a people-leader-first, technically-deep role. You are not expected to be writing production code day to day. You are expected to be technical enough to own the architecture alongside your team, review a design and find the thing that will break in the field, hold a line on a decision about memory, timing, or hardware interaction, and be trusted by your engineers as a technical peer rather than a coordinator.

Our product operates where embedded systems, machine learning, and hardware meet the real world. Decisions about memory, power, thermal budget, timing, and connectivity surface directly in what the customer experiences. There is no clean separation between the device and the product, and the constraints of the sensor are the constraints of the business. This is an on-site role with our in-person NYC team.

• The embedded and edge software platform: embedded Linux, the build and image toolchain, service orchestration, and the software running on every sensor we ship
• The hardware and software boundary: board bring-up, driver and interface work, and the path from a new hardware revision to shipping software
• On-sensor model deployment: the runtime, tooling, and pipeline that get our models onto constrained compute and keep them fast
• Field and deployment reliability: OTA updates, fleet health, and what actually happens on hardware nobody can physically touch
• The embedded roadmap and its execution, tied to annual company goals
• A team of embedded software engineers you will coach, develop, and hire

• Own the roadmap and delivery for embedded, and turn broad direction into clear objectives your team can execute against
• Hold the technical bar: review architecture, ask the question nobody else is asking, and surface the right tradeoffs to our Head of Engineering.
• Coach engineers on technical judgment and ownership, not just execution
• Hire and interview embedded and systems engineers as the team grows
• Partner with ML, RF, hardware, and mission operations so the device, the model, and the customer deployment stay in sync
• Own company goals and drive execution through ambiguity

• Software that behaves predictably on hardware in the field, under real power, thermal, and connectivity constraints
• A shorter, more repeatable path from a new hardware revision to shipping software
• Updates that reach deployed sensors without a person on site and without anyone…