models in training

Speech AI. Built small. Task trained.

Specialized speech models engineered for tight power and memory budgets. Recognition runs entirely on your hardware.

Request access Meet Corpus
01

Context shapes the model.
Vocabulary, acoustics and device limits.

Domain training

Specialized terminology and task-specific examples teach each model the distinctions its users depend on.

domain

Field acoustics

Wind, rotor wash, radio traffic. We train around the sounds that compete with a user's voice.

field

Device budgets

Compute, memory and battery limits guide model size and architecture from the start.

hardware
02

Efficiency at every layer.
Architecture, precision and runtime designed together.

Streaming recognition

Conformer architecture processes speech as it arrives, without waiting for a complete utterance.

arch

GPU-native kernels

Custom compute kernels target laptop and embedded GPUs, matching execution to the hardware.

kernels

1.58-bit weights

Targeting 30× lower power per parameter than FP16 with mixed 1-bit/1.58-bit weights. Benchmarks pending.

quant

Local deployment

Signed model files for offline inference. Audio stays on your device, with no telemetry or cloud dependency.

integrity
03

Corpus
Speech AI for casualty care.

in development

Hands on care.

A voice-operated, local-only guide to the Tactical Combat Casualty Care manual, with source-cited responses for trained responders.

Explore Corpus

Build with us.

We're seeking partners for early model evaluations.

Requests open soon. Your email has not been submitted.