Computational Imaging and Signal Processing Engineer

Merge Labs is a frontier research lab with the mission of bridging biological and artificial intelligence to maximize human ability, agency and experience. We’re pursuing this goal by developing fundamentally new approaches to brain-computer interfaces that interact with the brain at high bandwidth, integrate with advanced AI, and are ultimately safe and accessible for anyone to use.

We turn physical signals into information and back. We reason backward from neural readout and stimulation goals to co-optimize the physical and computational systems that make them possible. Working with scientists and engineers, we combine simulation, measurement, and signal processing across neural interfaces, device characterization, and high-throughput biological screening. Our systems stream tens of gigabits per second. Within constrained power envelopes, we find what limits sensitivity, resolution, and reliability, push what today’s devices can deliver, and set the specifications for future generations.

You will build and maintain a shared imaging core that turns raw ultrasound data into image series across our platforms. You will define performance objectives and develop, optimize, and rigorously validate reconstruction methods. A central challenge is making these methods fast enough for real-time imaging and efficient enough for large-scale offline processing. Early work includes bringing existing acquisition and processing pipelines into a common framework that other engineers can use and improve.

• Develop reusable methods for beamforming/back projection, clutter filtering, and motion and displacement estimation within a shared imaging core for processing and optimization.

• Optimize imaging pipelines through controlled experiments, balancing sensitivity, robustness, and computational cost.

• Rigorously validate imaging pipelines using reproducible benchmarks, reference measurements, and regression tests, quantifying performance and uncertainty even when ground truth is limited.

• Produce reliable imaging results, analyze them rigorously, and present findings concisely with their quality and limitations explicit.

• Meet sustained throughput and latency targets for real-time and large-scale offline processing within compute and memory budgets.

• Use measured imaging performance and system limits to shape device and acquisition requirements.

• Explore learned methods, including deep learning, and integrate them when independent evaluation shows a useful improvement over strong baselines.

• Strong signal-processing and computational-imaging fundamentals, with deep experience in at least one imaging or sensing modality.

• Experience developing or materially improving reconstruction methods that others use.

• Strong scientific Python skills and experience profiling and improving GPU performance and memory use for large datasets or high-rate data streams.

• Experience designing and maintaining shared scientific software, with clear interfaces, tests, and documentation.

• Sound judgment in experimental design, optimization, and interpreting measurements.

• A track record of rigorous validation and honest reporting of uncertainty, limitations, and negative findings.

• Experience turning ambiguous scientific needs into useful engineering results with colleagues across disciplines.

• Ultrasound or related wave-based imaging.

• Prior experience with real-time reconstruction.

• Learned reconstruction or denoising.

• Wave-propagation simulation.

• C++ or custom GPU-kernel development.

• Experience turning scientific Python prototypes into production-quality software, including work with build systems such as Bazel.

• Experience using AI tools or agents…