SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Transfyr is building physical AI systems for science—capturing real scientific work and converting it into high-fidelity, machine-readable records of experimental execution. The company addresses a critical gap: scientists lack the detailed feedback loop that athletes have (film room, instant replay). Instead, protocols describe intent, publications record lossy outcomes, and the small decisions that determine success or failure often vanish when work ends or scientists leave. This missing record is why automating physical science work has been so difficult, why training depends on scarce apprenticeships, why tech transfer is expensive and slow, and why scientists struggle to distinguish biological noise from process variability.
As a Forward Deployed Engineer, you will embed with scientists and partners in active lab environments to turn messy scientific data into actionable insights. This is not a conventional software-only role. You will collaborate with Field Application Engineers to ensure the full system delivers value. Some days you'll map protocols and interview experts; others you'll write code, analyze multimodal data, test product workflows, or troubleshoot deployments, hardware, and networking issues. The role demands wet-lab fluency, enough engineering skill to prototype, strong product judgment, and a deep bias toward learning from users.
Key responsibilities include: learning real workflows by embedding with scientists to understand experimental intent, execution, exceptions, and tacit knowledge; making science legible by turning observations, interviews, sensor data, and outcomes into structured representations; designing right-first solutions by building conviction before overbuilding; prototyping in the field by writing code, analyzing data, and creating lightweight workflows; supporting technical transfer as protocols move between experts and sites; translating field insights into product requirements; delivering customer value by owning whether deployments produce useful outcomes; building reusable FDE methods and tools; and laying groundwork for automation by converting execution-level data into troubleshooting agents and training systems.
You are high-agency, biased toward action, successful in ambiguity, customer-obsessed, and hands-on. You do not wait for perfect specs; you learn what matters, create missing structure, and move. You prototype quickly, test assumptions with users, and treat field failures as information. You can make progress when workflows change, data is incomplete, and customers don't yet know what to ask for. You earn trust by listening carefully, understanding the scientist's reality, and staying accountable to outcomes.
The role is in-person in Cambridge, MA, with regular travel to customer and partner sites (up to 20%). Transfyr is backed by a $25M seed round, collaborates with frontier AI labs, and has advisors including Chris Ré (Stanford), David Baker (Nobel laureate), Kevin Weil (fmr CPO at OpenAI), and others. The company is tackling hard problems at the intersection of frontier science, perception, machine learning, and robotics.
