Principal Embedded Software Engineer

Principal Embedded Software EngineerFluidic Controls & Instrumentation — BrainEx PlatformBexorg, Inc. On-site, New Haven, CTBexorg is on a mission to move medicine forward by developing technologies that improve and extend human life.As a fast-growing biotech company, we operate at the intersection of scientific rigor and startup velocity—where strong execution, innovation, collaboration, and accountability are critical to our success.Our work combines cutting-edge neuroscience, AI, and translational science to accelerate discovery and development for central nervous system (CNS) diseases.At Bexorg, we value individuals who are proactive, adaptable, intellectually curious, and energized by solving complex problems in a highly collaborative environment.The RoleWe are seeking a Principal Embedded Software Engineer to own the electronics, firmware, and real-time control that make the BrainEx whole-brain perfusion platform work. BrainEx is a proprietary closed-loop physiological system that functions as an artificial cardiovascular, pulmonary, and renal system for an intact human brain — custom hardware and software regulating circulatory volume, cerebral blood flow, systolic pressure and pulsatility, pH, dissolved gases, temperature, metabolites, and drug dosing throughout each run.This is deliberately not a pure-software job. We are looking for a genuine generalist — someone who writes production firmware in the morning, scopes a noisy pressure transducer signal after lunch, redesigns a fluidic manifold in CAD, and has a prototype printed or sent out for machining by the end of the week. You will be the person who takes a control problem from physiological requirement to schematic, to board, to firmware, to the Python supervisory layer above it, to a validated rig revision running on the production fleet.You will report into the Technology team and work shoulder-to-shoulder with bioprocess engineers, surgeons, perfusion technicians, software engineers, and drug discovery scientists. Current priorities include tightening closed-loop control, extending run duration (today up to 24 hours, with active work toward multi-day and ultimately two-week runs), improving reproducibility across donors, and expanding the experimental envelope the rig can support.What You'll OwnEmbedded Firmware & Real-Time ControlOwn the embedded stack on BrainEx rigs — bare-metal and RTOS firmware (C/C++ on ARM Cortex-M; FreeRTOS or Zephyr) plus embedded Linux supervisory layers — from architecture through field deployment and versioned rollout across the fleet.Implement the control loops that hold the brain in physiological range: PID, cascade, and feedforward control of pressure, flow, pulsatile waveform generation, temperature, pH, and gas exchange, with anti-windup, gain scheduling, bumpless transfer, and graceful degradation.Build the state machines and interlocks governing rig setup and priming, calibration, run initiation, steady-state operation, perturbation and dosing protocols, fault detection, and safe shutdown — including watchdogs, sensor-plausibility checks, and defined safe states for every credible failure mode.Guarantee determinism where it matters: loop-rate budgeting, interrupt and DMA design, jitter measurement, and instrumented timing so control behavior is provable rather than assumed.Own the field buses and device integration — I²C, SPI, UART, RS-485/Modbus RTU, CAN/CANopen, EtherCAT or equivalent, USB, and Ethernet — across pumps, gas mixers, mass flow controllers, proportional valves, dialysis modules, and analyzers.Instrumentation, Electronics & SensingDesign and bring up custom electronics: schematic capture and PCB layout (Altium, KiCad, or similar), analog front ends for bridge and ISE sensors, 16–24-bit ADC selection, isolation, EMI and grounding strategy, connectorization, and harnessing for a wet, salt-spray-adjacent lab environment.Integrate and characterize the sensor suite: pressure transducers, ultrasonic and Coriolis flow meters, thermocouples/RTDs, pH and blood-gas analyzers, inline optical O? and metabolite sensors, bubble detectors, and load cells for gravimetric fluid balance — including calibration procedures, drift compensation, and uncertainty budgets.Drive actuators precisely: stepper and BLDC motor control for peristaltic and centrifugal pumps, encoder feedback, solenoid and proportional valve drive, syringe and infusion pump control for compound dosing, and closed-loop verification of delivered volume rather than commanded volume.Debug at every layer — oscilloscope, logic analyzer, SWD/JTAG, protocol sniffers, and bench fluidic loops — and close the loop with design fixes that prevent recurrence across the deployed fleet.Mechanical Design & FabricationDesign fluidic and mechanical hardware in CAD (SolidWorks, Onshape, or similar): manifolds, sensor housings, reservoirs, bubble traps, sealed enclosures, and rig-integration hardware, with attention to O-ring and gasket sealing, dead-volume minimization, tolerance stack-ups, and biocompatible, cleanable material selection.Prototype fast and manufacture well: SLA and FDM 3D printing in-house for rapid fluidic iteration, and machined parts for production-grade hardware — you should design for the process, speak fluently with machinists, and be able to run the equipment yourself as we expand in-house precision machining capability.Select and qualify wetted components: tubing, connectors, quick-disconnects, and polymers evaluated for chemical compatibility with our synthetic-blood perfusate, sterilization or sanitization method, hemocompatibility, and long-duration performance.Software Integration, Telemetry & Operator ExperienceBuild the Python supervisory layer above the firmware: control services, device abstraction, protocol definitions, configuration management, and reproducible run scripting so that a protocol change is a versioned artifact rather than a knob turned by hand.Write the tooling that makes the hardware tractable: Python calibration routines, bench automation, sensor-characterization harnesses, log parsers, and analysis notebooks — the instrumentation around the instrument.Deliver real-time monitoring, alarming, and operator interfaces that let a perfusion technician supervise a multi-day run safely with minimal technical intervention, with a rational alarm hierarchy and clear operator guidance on every fault.Stream and structure the data: telemetry pipelines into our AWS data lake, integration with LIMS/ELN tooling and higher-level orchestration, and time-based alignment so physiological signals, dosing events, and sampling are analyzable downstream.Establish engineering hygiene appropriate to a mission-critical instrument: version control, code review, CI, hardware-in-the-loop and simulated-rig testing, firmware release process, traceability, and rollback.Reliability, Iteration & CollaborationMove fast without breaking runs: rapidly prototype mechanical, fluidic, electrical, firmware, and control changes, then bring them to production-grade rig revisions on a fast cadence.Define acceptance tests and performance metrics for each revision — control-loop performance envelopes, sensor accuracy, perfusion confirmation metrics, BBB integrity, run-success rate — and drive mean time between failures, run yield, reproducibility across donors, and the platform's duration ceiling.Serve as a senior technical authority on embedded and control architecture within Bexorg, setting standards and mentoring junior engineers, technicians, and perfusionists on the rig's design, operation, and troubleshooting.Translate experimental need into hardware: work directly with neuroscientists, surgeons, perfusion specialists, and drug discovery teams, and communicate clearly with non-technical stakeholders including collaborators at major biopharma companies, foundation partners, and government grantors.What We're Looking ForRequired QualificationsExtensive hands-on experience shipping embedded systems in production or mission-critical environments — instruments, medical devices, lab automation, industrial control, or comparable — where a firmware defect has physical consequences.Deep fluency in C/C++ on microcontrollers, with real working knowledge of RTOS concurrency, interrupt design, memory constraints, and low-level peripheral programming.Strong Python skills — not just scripting. You will build supervisory control services, device abstraction layers, calibration and test harnesses, data pipelines, and analysis tooling in Python, with the engineering discipline (typing, testing, packaging, async I/O, version control) that production instrument software requires.Demonstrated track record of owning projects end-to-end: requirements, schematic and board design, firmware, bring-up, commissioning, validation, and long-term support.Strong practical control-systems ability — you can tune a real loop on real hardware, reason about stability and sensor noise, and implement PID