Integrating AI into a Mobile App

Building artificial intelligence into an app for a London based design agency specialising in digital transformation, to deliver game-changing insights to users.

Integrating AI into a Mobile App

The Opportunity

Our client was working for a multi-national FMCG company with an annual turnover approaching 10 billion USD.

They had the expertise to build impressive digital products and needed a partner to deliver the AI component of the project.

We were required to build a component that could put AI in the hands of everyday customers, through surfacing smart insights from a variety of data sources. The module would need to integrate seamlessly into the existing application.

What we did

✔︎ Exploratory analysis of order log and point of sale data to establish insights that could be automated
✔︎ Built an engine to generate recommendations and handle edge cases smoothly
✔︎ Developed a Django application to handle the insight generation process
✔︎ Built a REST API to serve insights to the application
✔︎ Delivered a reporting framework to monitor usage and sales uplift for outlets using the application

The Results

A well as a measurable positive impact on sales, end users were able to provide feedback to the system regarding how useful and actionable each insight was to their business, completing an intelligent feedback loop.

The application was built to the agency’s own high standards of code quality and the close working relationship throughout the build ensured the handover was seamless, as the team already had a full knowledge of the deliverables.

7%

increase in sales from users using the app and interacting with the AI generated insights

How we did it

We built an independent AI module, designed to integrate smoothly with the core application that the agency was building.

The module scans streams of internal and third-party data for insights and surfaces the most relevant to the app, using a smart, personalised scoring system designed to highlight actionable insights first. The Python application was deployed in Docker with an Django API interface.

The success of the project has been amplified by the tight integration between our data science team and the wider project team. We remained aligned with the in-house designers and developers through a close partnership that ensures the AI component is not a 'black-box', but instead is a well understood, interpretable component of the app.

Our emphasis on practicality over unnecessary complexity has ensured that the insights are easily actionable by end-users.

Looking to run a similar project?

If you’re interested in finding out more about our services and how they can transform your business, get in touch and we'd be happy to tell you more.

Sales Propensity Model
Private Equity: Driving Value from Data