HVNT, a pioneering Augmented Reality games company, has developed an AR mobile game which turns cities into location-based treasure hunts. The game involves following clues, typically images, text or stylised treasure maps, to reach hidden locations in the city. Initially, HVNT were writing these clues manually, meaning that creating treasure hunts for new locations was an extensive process. Datatonic worked with HVNT to implement an automated system by exploring the application of Machine Learning models to generate clues, reducing the manual effort required by content creators, and enabling HVNT to expand their game more rapidly.
Automated clue generation for HVNT, greatly reducing manual efforts in creating clues.
Implemented solutions to accelerate the geographical growth of the game, creating the potential to have an increased number of players.
Enabled HVNT to provide their users with more clues and locations, meaning they can focus their efforts on expanding into more cities at a faster rate.
HVNT is a startup based in Stockholm, Sweden that has developed an Augmented Reality mobile game, HVNT, which turns cities into location-based treasure hunts. The game involves following clues to reach hidden locations in the city. These locations are usually landmarks such as parks and tourist attractions, and the clues are typically images, text or stylised treasure maps, based on the locations where they are found.
Before working with Datatonic, HVNT was finding it difficult to scale up the creation of clues that are required to make new treasure hunts. HVNT’s clues had to be written manually by their team of content creators, which was very time-consuming.
To address this, HVNT had already built out a Firebase web application which allowed their team to manually create clues for a specific location. However, this meant that in order to be able to create clues, the content creators needed to access data about the locations. The Firebase application that HVNT was using would trigger a Cloud Run instance to get raw Wikipedia data by calling the Wikipedia API. This added some friction to the process, as clues from raw data often had to be edited after the information had been extracted.
“Until now, HVNT had a group of content creators who would manually write the clues for each of these locations.”
Datatonic worked with HVNT to develop a reliable system to automatically generate clues for the treasure hunts in their game. The main goals of this project were to reduce the manual effort required by the content creators, and to allow for increased growth of the game.
While working on this project, Datatonic worked closely alongside HVNT to:
With our solution in place, HVNT are able to generate clues a lot faster, allowing them to scale up their game with much less manual effort. They are also able to provide their users with more clues and locations, and can focus their efforts on expanding into more cities at a faster rate.
“Datatonic helped us build a flexible cloud solution for automatically generating location-based content for our game. This significantly improves our abilities to generate fun and exciting clues for our users and to do so in a more efficient manner.”
If there’s data, we can optimise, innovate with and automate it to drive business value. Find out how we’ve supported the biggest brands across industries.
Know exactly where and how to start your AI journey with Datatonic’s
three-week AI Innovation Jumpstart *.
* Duration dependent on data complexity and use case chosen for POC model
With your own data sets, convince your business of the value of migrating your data warehouse, data lake and/or streaming platform to the cloud in four weeks.
With your own data, see how Looker can modernise your BI needs
with Datatonic’s two-week Showcase.