An innovative social startup partnered with CoreValue’s “Core Mobile” team to create a “Hyper Local Social Platform” to connect like minded individuals at the right locations at the right times and ensure the best night ever! It provides a clean & easy to use mobile app delivering venue information, ratings, food and drink choices. The app’s creators wanted to go further and help insure its users would have a great night by providing insights including:

    • Male / Female ratio
    • Average age ratio
    • Complete guest trend analysis
    • % venue capacity

To navigate through the oceans of data needed and provide the insights the start-up was looking for CoreValue’s “Core Intelligence” team to be called into action.


The Core Intelligence team was faced with a number of challenges to provide critical insights needed for the Application. To seed a representative dataset needed for new user adoption, alternative data feeds and social platforms had to be leveraged. Proximity data needed to be analyzed to focus processing requirements and minimize the potential for excessive compute costs. Multiple social media platforms needed to be analyzed to determine venue sentiment and predict gender of venue guests. Ultimately, this analysis was critical to the adoption and success of the application.


To deliver the key insights required, the Core Intelligence team quickly engaged and worked closely with the customer. The team realized it’s critical to understand the business challenge and context before starting to look for a solution. The team also worked closely with the Customer’s team to understand the data sources that were available and to get an initial idea of the quality and completeness of this data; a critical task for further analysis. After a deep exploratory analysis of the data, the team documented their initial findings and reviewed with the Customer’s team. The team proposed three key areas for initial focus and began initial modeling.

  • Gender prediction algorithm based on twitter name
  • Spatial Optimization Algorithm
  • Model to identify sentiment of the text from blogs & twitter feeds

After initial success the models were tuned and eventually deployed into production. The team used R for generating the models and opted to leverage MongoDB based on the data size and scalability requirements. As shown in the image below, models were reviewed and dynamically tested by the extended team.

casestudy2 img - Case study two


The applications were launched on both Google Play and the Apple App Store providing the initial experience and answers Customer’s users were after. In addition Venue owners are able to participate in the community and get a much richer understanding of their customers behaviors. By having aggregated reviews, trend analysis, check in data, business owners are able to understand and improve customer satisfaction, retention and lifetime value. The app we developed now provides a wealth of opportunities to both guests and business venue owners.