The healthcare field needs AI tools for two very different families of applications: insurance and clinical (med care). The objective is to build AI tools that can assist with the “knowledge work” of healthcare professionals on both sides. These include patients (for self-care), nurses, case managers and medical assistants, medical doctors (MDs), therapists (e.g., psychological, CBT, etc.), medical call-center support agents and insurance claim adjusters.
These groups have different needs, described below.
- For patients → e-health “call-centers” have limited scripts available during a call. Agents put patients on hold for a few minutes and go do manual search from the database for unexpected queries. HIVE™ can help with search-retrieval-answering for such queries in real-time.
- Nurses, case managers and medical assistants take perfect notes during meetings, patient handovers as well as patient interactions. However, naturally they can miss important information (e.g., “patient vomits so you cannot prescribe pills”) during these interactions. HIVE™ can help them by summarizing all important info if the transcript for the interaction / call / meeting is available.
- If the full transcript is not available, and/or if it’s a live interaction (e.g., logging intake info for a returning patient at the ER) then HIVE™ can help by showing suggested questions, answering live questions from the professionals (like in the case of live support agents), or giving relevant summaries from previous interactions in real-time.
- Insurance claim adjusters are on the other side. They have to validate insurance claims for patients and make sure regulations are followed. For this, they typically need a tool that can analyze claims and confirm cases. For instance, the adjuster might look at the treatment plan/logs and ask “is this procedure consistent with the AMA guidelines?”. HIVE™ can help the adjuster by pulling relevant information from the guidelines and comparing that with the patient records.
