Elomia Health has acquired Earkick, according to Tech.eu. Earkick is an AI companion that more than 250,000 people use to vent and to follow patterns in mood, sleep, activity and daily habits. The source does not give the price or terms of the deal.
What Elomia gets
The deal brings Earkick's AI models into Elomia's platform. It also gives Elomia ongoing early access to new versions of Mia21, a memory engine that the Earkick team originally built. Elomia's founder, Taras Pohrebniak, says the engine tracks a person's state and context over time. He says it will be combined with Elomia's own models and its safety and overreliance-prevention layer, and become a core part of the technology.
The strategic shift
Elomia was founded in 2019 as an R&D lab studying how large language models can support wellbeing. The company launched with a consumer app of its own. It is now moving into offering its API to organisations focused on care, and it says its aim is to become the foremost underlying platform for AI-based wellbeing support. The API lets organisations such as healthcare providers launch chatbots for wellbeing support that are built to be safe and compliance-ready. The article says Elomia's technology has been studied independently, including in a recent randomised controlled trial with the University of Pennsylvania.
Earkick was founded in 2021 by Dr Herbert Bay, co-inventor of the SURF computer vision algorithm, and Karin Andrea Stephan. The app gives personalised suggestions and content based on a user's own data. Stephan, now Earkick's COO, argues that the right message at the right time can save lives and that the combination makes such support available at scale.
What is not in the source
The article gives no integration timeline, no pricing for the API and no detail on how Earkick's consumer app will continue. Those are open questions for anyone evaluating the platform.
Why it matters beyond healthcare
The deal shows a pattern: a consumer AI app is folded into a vendor that sells the same capability as an API to organisations. Persistent memory of a user's context and a layer that limits over-reliance are the parts being consolidated. Any team that puts a conversational bot in front of customers will face the same two questions about memory and limits.