Kintsugi Voice is an API-first platform, we are developing novel voice biomarker infrastructure to detect signs of clinical depression and anxiety from short clips of free form speech. We help healthcare organizations scale access to mental healthcare with smarter, more accurate triage. Our technology supports multiple use-cases to help more people get the right care at the right time.  We focus on early detection that allows you to lower cost and improve patient outcomes. We have created the first clinical-grade voice biomarker diagnostic (U.S. Patent-Pending and FDA De Novo pending) to measure the severity of depression and anxiety in real-time.

Brim Analytics is driving to the next cure, faster with a democratized AI-guided chart abstraction system. Chart abstraction is the process of converting unstructured clinical notes into structured data, an essential task across healthcare, including clinical trial matching, academic research, registries for cancer and heart disease, and many more. Brim is an AI-guided tool for chart abstraction that prioritizes privacy and security, creates traceability of data, and is dramatically faster than human curators alone.

Faster, smarter chart abstraction to streamline clinical workflows, pre-screen clinical trials, accelerate retrospective research, and more.

1. Kintsugi Voice only needs 20 seconds of free from speech vs. a manual screen that takes roughly 10 minutes.
2. Kintsugi Voice is language agnostic and does not listen to what a patient is saying, but how they are saying it.
3. Kintsugi has the largest annotated patient data model, validated by a psychiatrist team with more than 7 years experience. This allows us to predict depression and anxiety with high specificity and severity, based off of PHQ-9 standards.
4. Kintsugi Voice’s flexible API enables hundreds of workflows from retrospective to live in-call scoring. Typical depression screenings are typically delivered 1 time and not longitudinal. We provide the ability to longitudinally track patients across a period of time.

1. We are not a NLP tool, we do not need any canned speech and or paragraph read to detect depression and or anxiety.
2. We have the largest annotated data model in the world with over 35k patients.
3. We are supporting over 1M at risk lives.
4. 2X accuracy rating compared to the standard of care.

Brim enables fast and flexible abstraction of structured data from unstructured notes.

• Fast Abstraction. AI-guided generation saves 80% of the time relative to manual abstraction with comparable accuracy.
• Scalable. Abstract data at scale, rather than note-by-note.
• Flexible. Easily iterate on the fields you derive as you learn or your process evolves.
• EHR-agnostic. Import and export via a standard format.

• Privacy and Security first. Brim is fully secure, and we do not train on your data. You can implement in your institution's cloud environment, or in a single-tenant deployment. Leverage your institution's LLM endpoint. Brim supports Open AI, Azure Open AI, Bedrock, and other LLMs to get the newest foundational models without adding another vendor.
• Built for researchers and abstractors. Purpose-built interface for abstraction helps humans move faster and improves the accuracy of abstracted data.
• Domain flexibility. Brim has been used for projects in cancer research, clinical trials, orthopedic injuries, and many more. Brim optimizes variables for your project based on your feedback, improving abstraction in real time.