Technology Notes: Advancing Our Personal Care Agents and dHealth Intelligence Platform

Tech Update #1 - October 2026

Over the past year, our team made dHealth Intelligence available through a subscription for AIDH holders - and this layer is being extended to provide end-to-end encrypted memory storage.

In fact, healthcare imposes some rules that we can't just ignore; therefore, any dataset that will be stored in your client storage will also be encrypted to preserve privacy.

Next to the dHealth Intelligence platform, we have started to build a console for you and/or healthcare professionals to create and customize Personal Health Agents and Care Partner Agents, using the Practitioner Console demo.

On that same basis, we have also released the Care Partner for Dementia Caregivers and family members of patients suffering from dementia. Fill in some information about the patient, and an AI agent will assist you with education, orientation and some everyday relief for coping with the disease. Of course, this is an education tool, not a medical device. Our clinical study at Bareilly International University with families of patients suffering from heart failure applies the same logic.

Following this principle of creating scoped forms that permit the creation of Personal Health Agents and Care Partner Agents, we also released the Prevention Agent, which you could even use yourself if you want to reduce stress or improve your sleep.

And this is just the beginning of how dHealth Intelligence may be integrated by Healthcare practitioners or even directly by the community. What is important to note for these demos and forms is that we do not store any single dataset related to these forms responses on our servers - and the configuration of AI agents always sits in the generated links and QR codes, which healthcare practitioners may share with their patients over any preferred communication channels - Mail, Whatsapp, etc. 

It’s a rapid form of integration - [pun slightly intended] - that gets AI agents directly in the patient’s hands, i.e. by a single click on a link or by scanning the QR Code received from a Healthcare specialist. No extra setup necessary. And it respects your privacy.

We are also internally conducting several experiments related to hardware devices, and edge devices that may be used to host our Personal Care Agents entirely - whilst it is already feasible today to run smaller language models (SLMs) on edge devices, it is increasingly more costly to run higher-level models (VLM / LLM).

But as the field moves rapidly towards more effective models, we believe that for some use cases, some tools and devices may be used to self-host your Personal Care Agents as well.

—

That's about the tech that runs dHealth Intelligence. There are many emerging use cases for Healthcare with AI Agents, but as I said in the opening, we have to move cautiously and use state-of-the-art software to keep patients' data private. Several partnerships have been announced where our rapid integration will hopefully get us some more usage, too. I don’t have numbers now, and I will make sure that next time, my monthly updates are worked on during the entire month and contain some business partners' information as well.

—

What does your Personal Care Agent need to handle for you?

As usual, feel free to comment on this post with your answers.

You may also contact us with ideas or feedback and inquiries through our Contact Page.