Case Study

Making better decisions

How AI improved personalised healthcare using genetic data

If you identify a gap in your services, could AI be used to speed up the decision making process on what a potential solution or product could look like? That’s exactly what we helped our healthcare company client to achieve.

The client currently offered a genetic testing service. The test gave their customers a detailed insight into their own health. DNA testing reveals a wealth of information for individuals, like which foods or medicines they may be sensitive to, and how susceptible they are to disease.

The testing service also included specific recommendations on lifestyle for customers, advising how to mitigate their unique risk factors, and improve their overall health.

The solution

We worked in collaboration with our partner The Startup Factory on a Design Sprint – a week long structured process – to establish how an AI solution could help solve these problems, and to quickly build a working prototype of an AI-enabled mobile app. (You can read more about the Design Sprint process in our blog here)

The Design Sprint meant the client was in a position to make faster, informed decisions around designing the product to take to market, and be confident in their choices.

The result of this process, the prototype app, provided dietary recommendations to users based on the results of their DNA testing, collecting feedback from the user on the quality of the recommendations, and qualitative information on how the user felt at the end of each day.

The process gave us an in-depth understanding of what was important to both client and customer, and allowed us to explore the ways in which AI could be useful. For example:

  • Using context to target better recommendations: The user’s daily routine, location and habits could be considered in order to deliver better individual recommendations.
  • Better targeting through segmentation: Segmentation algorithms could help discover groups of similar users based on behaviours and preferences.
  • Integrating recommendations into other devices: For example, using Fitbit data (or similar devices) to provide relevant exercise recommendations, and help users track progress against their goals. Or using food label barcode scans on a smartphone to inform diet recommendations.
  • Learning from user behaviour and feedback: Understanding which recommendations were effective and which were not generates a labelled dataset. Machine learning algorithms could use this ‘what happened next’ data to improve the quality of recommendations to users.

With our expertise and input, we were able to help fully map out the journey from problem to solution for our client. We deployed the right AI technology to build a working prototype, giving the client the confidence to develop the right product to take to market.

Our other case studies: