Get our free monthly newsletter
The latest news, case studies, events & opportunities across the creative industries.
Thank you! You are now subscribed to our newsletter.
Oops! Something went wrong while submitting the form.

By clicking the Join Now button, you agree to our Terms of Service and Privacy Policy.

Cookies Preferences
Close Cookie Preference Manager
Cookie Settings
By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage and assist in our marketing efforts. More info
Strictly Necessary (Always Active)
Cookies required to enable basic website functionality.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Ones To Watch 2021

In game development, the game AI is what controls the behaviour of every single entity. It could be a bot playing your enemy, the final boss, or a character that plays alongside you. These game AIs are in a vast majority scripted, they follow a specific set of rules and it requires a huge amount of work to turn them into believable characters. While game AIs haven't changed for years, the complexity of the games we play has increased over the past years (procedural generation, emerging gameplays that let you create your own world i.e. Minecraft, Roblox) and it is getting very difficult to develop scripted bots for certain games.
That’s why we created WildMeta in late 2019, to help video game developers build smarter bots and non-playable entities using machine learning. After months of bootstrapping our core technology, we have created a framework that enables us to train machine learning AIs to play a video game as a bot, NPC (non-playable character) or player.
Thanks to their neural network, our bots play the game they've been trained on. They interact naturally with their environment and react to the player's moves which improves players’ experience and keeps them engaged in a game. While we believe machine learning will play a major role in the future of game development, we don't expect most video studios to hire a team of machine learning experts anytime soon, that's why we are offering our expertise and technology as a service. We act as studios' specialised outsourcing partner so game developers don't need to understand machine learning to have smarter bots. This way we can look after the game AIs and their integration, and studios can focus on their game.

My co-founder Alexandre Borghi, PhD, and I have known each other for 7 years. We started WildMeta after closely following the work done by academic researchers using video games to train machine learning agents. But instead of trying to beat every single human player at a game, we saw a huge potential to use this technology to improve players' experience and support game developers!

We have managed to bootstrap the development of cutting edge technology. We have turned academic research into a solution ready for production with no external funding. We have no doubt our work can impact the entire game industry, but to drive the adoption of such new technology, we need first to address game studios' assumptions when it comes to machine learning. We were initially building a B2C platform directly for players and pivoted to a B2B service-based model in January 2021. Since then, we have been more public about our existing technology, gave public talks (Pocket Gamer Connect, April 2021), were featured in several podcasts and shared blog posts.Today we are working on our next demonstration and talking to potential partners for 2021-2022 projects.


In the years to come, we want to play a major role in bringing machine learning to game development and I believe we are in a perfect position to do so! Studios are familiar with the outsourcing approach and understand how this model can help them save time and money. They usually have a dedicated budget tight to a single game which means it is difficult for them to justify the cost of an R&D team to build from scratch machine learning systems. The good news is, we've already done all the R&D so studios don't need to, which also means we can start working on their game straight away. As soon as more studios lead the way, I have no doubt it will become an industry standard that will greatly improve players' in-game experience.

How being a CIC CreaTech One to Watch 2021 will help us? Well, we've been talking to a number of studios and while they are genuinely curious about this technology, there are still many assumptions to address on machine learning and how it fits into their game, that is the reason why I am dedicating some of my time to give presentations at industry events and open discussions with game developers. I believe being a CIC CreaTech One to Watch 2021 would help us spread the word, open new doors with studios that may have been looking into using machine learning for their game but didn't know where to start, and help us drive adoption in the UK game community.

View More

designerzen ltd

KAVE Immersive Ltd

Limbic Cinema

Data Duopoly