IHCP Platform
The clinical layer that listens for what language does not say
IHCP is not a translation tool. It is an interpretive intelligence layer that detects emotional signals, cognitive state, and cultural context beneath language - surfacing what a patient is actually experiencing, not just what language they are speaking.
The Problem It Solves
Standard translation tools address words. They were built for transactional exchanges, not clinical ones. They assume a single language pair. They cannot detect affective or cognitive signal. They were not designed for care workflows.
What they miss is the layer underneath: a patient who is frightened but not saying so. Agitation that is building. A care interaction where the language the patient is speaking right now is not the language noted in their file - because cognition has shifted, and the language they learned first is the one that remains.
That is the gap IHCP was built to close.
What IHCP Does
Real-Time Affective Signal Detection Surfaces emotional state, distress, and agitation as percentage-based outputs - designed to support clinical judgment, not replace it.
Multilingual Oscillation Handling Supports natural movement between languages, dialects, and cultural frames within a single interaction, without assuming a fixed language pair.
Audio to Audio Real-time bidirectional communication calibrated to patient uptake speed. Emotional and cognitive signal lives in voice, not text.
Cultural Context Interpretation Goes beyond words to the meaning underneath. Built for populations where care interactions carry cultural context that generic tools cannot parse.
Sovereign Architecture No third-party translation APIs in the clinical workflow. No patient data retained. 100% Canadian-owned IP. PHI-compliant by design.
Model as a Service Deploys embedded into partner applications, devices, or care kits. The model loads directly into the client's environment - data stays local and never leaves their infrastructure.
Products Built on IHCP
TalkLikeMe - Voice and cultural communication for senior care
Designed for long-term care environments where multilingual oscillation creates real clinical risk. As dementia progresses, patients frequently lose access to later-acquired languages and revert to their first language or childhood version of it. In many care environments, no common language exists between patient and staff at all.
TalkLikeMe was built with Dania Care Society in BC, co-developed with clinical staff, and validated against clinical ground truth on first use. It supports care teams and residents across a range of settings and interaction types - with on-device processing and no patient data stored.
IP Ownership: 50/50 with Dania Care Society
Deployment: Live at Dania Care Society, BC
MIRA - Clinical affective signal detection, any setting
The broader intelligence layer. Sentiment analysis, agitation detection, and emotional signal surfacing across languages and care contexts. Deployable across a wide range of environments: integrated into existing clinical infrastructure, embedded, or used as a foundation for custom builds.
MIRA extends IHCP's core capabilities into general healthcare settings - prenatal care, women's health, acute care, in-home care - anywhere that communication gaps create clinical risk.
IP Ownership: 100% HM Technologies
Status: In active deployment discussions across multiple care settings
Why It Is Different
Language drift as cognition changes
Multilingual oscillation within a single interaction
Affective and cognitive signal beneath the words
Deployment without third-party APIs or data leaving the care environment
Most platforms that address language in healthcare were built for transactional exchanges and adapted for clinical use. IHCP was built from the ground up for care environments where communication failure is most consequential. It handles what others cannot:
Proof
Dania Care Society selected TalkLikeMe after evaluating four commercial alternatives. It was not an informal decision: it was a structured comparative assessment by clinical professionals who chose IHCP on clinical fit and operational performance.
During the first week of live deployment, a resident communicated severe abdominal pain that had gone undetected for four days. A human interpreter arriving independently that same morning confirmed the same finding. Clinical ground truth, validated in real time.
The platform has since been presented at the International Society of Gerontechnology 2026 conference, including a special invite-only session at the Dania deployment site. International partnerships are in active negotiation in Southeast Asia and Africa.
If You Want to Build on IHCP
The model deploys as a model-as-a-service. Front-ends are optional. If you are a care organization, hardware vendor, or technology partner working on a problem that needs this layer, that is a conversation we welcome.
→ Book a Conversation → Request a Demo