Voice and language technology built around your requirements.
Cobalt develops speech recognition, voice intelligence, natural language and conversational interface technology that can run on-premise, in a private cloud, or directly on your device.
How We Help the Speech-Language World
We partner with innovators across the speech-language ecosystem to develop technology that improves clinical workflows and language assessment.
- Speech and language technology
- Custom software development
- AI-powered language tools
- Technology partnerships
Partner With Cobalt
Whether you're looking to build, enhance, or integrate voice and language technology, our team is ready to collaborate on your next project.
Custom Solutions. Built for You.
We provide consulting and custom technology services to help organizations bring voice and language solutions to life.
- Consulting & technical expertise
- Custom product development
- Technology partnerships
- End-to-end implementation
SALT Software is Now Part of Cobalt
SALT Software is the leading platform for language sample analysis, helping speech-language pathologists collect, analyze, and interpret spoken language for assessment, treatment planning, progress monitoring, and research.
The SALT team has joined Cobalt. Together, we're continuing to develop new technologies that advance language assessment while supporting clinicians, researchers, universities, and healthcare organizations.
- Language sample analysis
Collect and analyze natural speech and language samples.
- Clinical assessment tools
Powerful tools that support accurate assessment and documentation.
- Research & education
Advancing language science through research and collaboration.
- Powered by Cobalt technology
Built on secure, accurate, and scalable technology.
Speech technology without the usual constraints.
Cobalt helps organizations deploy speech and language technology around their actual operating, security, infrastructure and domain requirements, not around a fixed public API.
Privacy Built-In
Keep sensitive audio, transcripts and voice data within the infrastructure your organization chooses.
Built to adapt
Customize acoustic and language models for specialized vocabulary, accents, acoustic environments, devices and workflows.
Flexible deployment
Embedded on a device, on-premise, in a private cloud, or as batch or streaming services.
A complete speech and language technology portfolio.
Three product families cover recognition, intelligence and interaction — with a research group for problems that off-the-shelf models do not address.
Speech Recognition & Transcription
Turn spoken language into accurate, usable text with models adapted to specialized environments and terminology.
Explore Speech productsVoice Intelligence
Protect sensitive information, authenticate speakers, and analyze characteristics contained in voice.
Explore Voice productsVoice User Interfaces
Build natural spoken interactions for applications, devices, portals, equipment and support systems.
Explore Voice productsDesigned for your environment.
- STEP 01
Understand the requirement
We start with the operating environment, security posture and use case.
- STEP 02
Select and configure
Choose the products and configurations that fit the deployment.
- STEP 03
Adapt models and interfaces
Tune acoustic, language and interaction models to the domain.
- STEP 04
Deploy in your infrastructure
Ship embedded, on-premise, in a private cloud, or as a service.
Speech and language technology for complex environments.
From customer conversations to clinical assessments, we deliver secure, accurate voice and language solutions that power critical work.
Contact Centers
Secure transcription, redaction, speaker verification and analytics for high-volume voice channels.
- –Batch and streaming call transcription
- –PII and protected-data redaction
AI Product Development
Custom voice and language technologies for organizations building next-generation products.
- –Custom software development
- –Voice AI integration
Healthcare & Medical Devices
Speech technology for clinical environments, medical devices and hospital operations.
- –Hands-free operation of medical equipment
- –Clinical vocabulary recognition
Speech-Language Pathology
Supporting clinicians, researchers, universities, and schools with language technology and assessment tools.
- –Language sample analysis
- –Clinical workflows & reporting
Research & Universities
Advanced speech and language tools that accelerate research and support education.
- –Research collaboration
- –Data collection & analysis
Your data stays under your control.
Cobalt technology can operate without transferring sensitive voice data to a public, multi-tenant speech platform. Deployment options are chosen to fit the organization's security posture and infrastructure.
When an off-the-shelf model isn't enough.
Cobalt adapts or develops speech and language systems for problems that general-purpose models handle poorly. Our team works across acoustic modeling, language modeling, NLU, TTS and speaker biometrics.
- Domain-specific terminology
- Noisy acoustic environments
- Alternate microphone configurations
- Accented speech
- Children's speech
- Dysarthric speech
- Specialized applications
- Embedded and edge environments
Where Cobalt technology is used.
Secure transcription for regulated industries
Batch and streaming ASR deployed inside the customer's infrastructure, so audio and transcripts never leave it.
Voice interfaces for self-service
Conversational voice interfaces built with Cobalt Interact for portals, applications and support systems.
Custom voice and biometric technology
Adapted models for speaker verification and identification in devices and applications.
Contact-center speech intelligence
Redaction, diarization and analytics for high-volume call audio.
What partners say about working with Cobalt.
“Cobalt Speech and Language is the perfect partner for a small business like ours to tackle the challenges of speech recognition. We have very much enjoyed working together.”
“We have been working with the Cobalt team on advanced methods of using the Vocalzoom HMC sensor for improving performance of existing speech recognition systems and for user authentication. Their extensive knowledge and capabilities allow us to innovate around the unique features of our sensor.”
“As an AI-powered solution in search of a speech application, Cobalt's ability to provide agile, custom speech solutions that scale on demand sets them apart from others in the market.”
“Jeff Adams and his team have contributed immensely to the success of FluentWorlds. They are both brilliant technologists and incredibly responsive. Our company would not be where it is today but for the outstanding contributions of the Cobalt team.”
Latest from Cobalt.
SALT Software Joins Cobalt Speech and Language to Pioneer the Next Generation of Clinical Assessment
The partnership combines 40+ years of gold-standard clinical evaluation and data with cutting-edge AI to automate assessment and empower speech-language pathologists.
Read more →Cobalt Speech Announces End-to-End Speech Recognition and Transcription Service
The leap in accuracy of automatic speech recognition (ASR) is largely due to the ongoing switch from a hybrid approach of carefully strung-together models to end-to-end (E2E) model architectures.
Read more →Cobalt Extends Application Window for Free Trial of Speech Recognition Technology
After a strong response to the initial free trial of Cobalt Transcribe, the application window was extended to give more companies and academic institutions time to apply.
Read more →Cobalt Speech and Language Offers Free Trial of Speech Recognition Technology
Cobalt invited companies and academic institutions to apply for a 100-day free trial of Cobalt Transcribe, its automatic speech recognition service.
Read more →Let's build speech technology around your requirements.
Talk with Cobalt about your application, infrastructure, privacy requirements and deployment environment.