Production AI applications
We build and ship end-user AI applications, including FVrad and Field Viewers Mammography AI, to close the loop from model output to operational workflow.
What we do
AI-first capabilities across a vertically integrated stack
Field Viewers leads with AI outcomes, then engineers every required layer to make those outcomes reliable: sensing physics, mixed-signal ICs, FPGA/DAQ, data infrastructure, model operations, and deployed software.
AI applications and intelligence layer
MLOps and deployment
Training, validation, inference serving, drift monitoring, and release workflows that move AI from prototype to accountable production operation.
Decision-grade analytics
Imaging and waveform intelligence pipelines tuned for interpretability, latency, and deployment constraints in regulated and mission-critical environments.
Models are only as stable as the signals feeding them. We ship production AI—FVrad and mammography products among them—with release discipline learned from live users, not Kaggle leaderboards.
Read the brief for how we tie MLOps, monitoring, and firmware-level feature contracts together when your program spans silicon and cloud.
Medical imaging & molecular imaging
Photon-counting CT
Detector modules, readout architecture, and data pathways toward energy-resolved CT concepts — partnering on channel count, rate, and power targets appropriate to your roadmap.
SPECT & PET
Detector interfacing, timing, coincidence logic support, and integration with reconstruction pipelines for nuclear medicine modalities.
Algorithms & reconstruction
Iterative and analytical reconstruction, corrections, simulation-driven development, and GPU/CPU implementation support.
Clinical and preclinical imaging programs need detector contacts, readout, and reconstruction to agree on a single noise budget. We work photon-counting CT, SPECT, and PET threads end to end.
The long-form brief covers AZO contact strategies for detector lifetime and charge collection, plus how those choices propagate into algorithms and AI.
Radiation detection & security
Radioisotope identification
Spectroscopy front-ends, stabilization, library matching hooks, and firmware that turns raw pulses into actionable identification features.
Real-time radiation dosimetry
Readout electronics and software for radiation dosimeters with live dose updates — automation, calibration, and field-ready workflows.
CBRNE program support
Hardened sensing subsystems, DAQ, and analysis stacks scoped to customer requirements for CBRNE scenarios.
Isotope ID, live dosimetry, and CBRNE architectures depend on front-ends and firmware that stay calibrated when the environment does not cooperate.
The capability brief also summarizes our U.S. Army SBIR Phase I award context—external validation of the detector and readout direction we bring to defense-relevant programs.
Hardware & integration
Capability brief: Hardware integration & detector electronics →
Semiconductor sensors & materials
CZT, CdTe, and related pixelated detectors; hybridization and characterization support alongside readout design.
Readout ASICs
Multichannel ASIC definition, bench validation, and integration with FPGA and host software — including high multiplexing ratios where programs demand it.
FPGA & DAQ
Real-time DSP, trigger logic, high-speed links, embedded control, and laboratory or OEM-grade DAQ products.
Packaging & SiP
System-in-package paths, wirebond and advanced packaging planning, and miniaturization for aerospace and portable instruments.
Multichannel ASICs, wirebond and hybrid integration, FPGA/DAQ, and detector contacts—including aluminum-doped zinc oxide (AZO) approaches for lifetime and charge collection—live in one engineering thread here.
The brief ties those threads to our patent portfolio positioning, DOE Phase I funding, and how we engage under NDA for freedom-to-operate discussions.
Software, AI/ML & digital health
AI / ML
Model design, training pipelines, deployment to edge or cloud, and validation planning for imaging and signal workloads.
Telehealth & RPM
Connected device software, secure data paths, and integrations that support telehealth and remote patient monitoring programs — engineering-focused, not direct patient care.
Remote diagnosis support
Imaging viewers, workflow tools, and infrastructure that help clinicians and physicists collaborate remotely; all deployments follow your regulatory strategy.
Digital health is not a separate skunkworks: our software patterns inherit from the same secure pipelines and logging discipline we use in shipped AI products.
Open the brief for telehealth/RPM integration, imaging viewers, and how we keep regulatory boundaries explicit in scope documents.
Specific claims, timelines, and regulatory classifications depend on the program. Share your use case on the contact page or use the on-site assistant for a structured overview (it is not a substitute for a formal statement of work).