Inworld has launched Runtime for developers, an AI toolkit that was born from experiences with industry giants like Disney, Xbox, and NVIDIA.
Every developer in the AI space knows the feeling. You’ve built a brilliant prototype, a glimpse into a future of truly interactive and engaging consumer experiences. It wows everyone who sees it, but then comes the hard part: how do you get it into the hands of a million users without it falling over or costing a fortune?
This is the scaling wall where so many great ideas get stuck, but it’s a problem Inworld AI believes it has cracked.
Inworld’s journey started with helping developers create immersive AI characters for games and interactive media. To do this for some of the biggest names in entertainment, they needed an infrastructure that was rock-solid, lightning-fast, and wouldn’t break the bank, even with millions of people chatting to their AI characters at once. They had to build it themselves because nothing else on the market could do the job.
Kylan Gibbs, CEO of Inworld AI, said: “Existing tools couldn’t deliver at the speed and scale our partners required. When we realised every consumer AI company faces these same barriers, we knew we had to open up what we’d built.
“We’ve watched the industry reach an inflection point. Thousands of builders are hitting the same scaling wall we did, so we hustled over the past year to add capabilities beyond our internal needs to create the universal backend to accelerate the entire consumer AI ecosystem.”
Through its work, the Inworld team and its developers learned that what separates a runaway success from a forgotten prototype comes down to a few hard-won lessons:
- Bridging the gap between a cool demo and a production-ready app, a process that can burn through six months and countless engineering hours.
- Freeing your best people from the endless cycle of maintenance and bug fixes so they have the creative energy to innovate.
- Being able to experiment and learn what users want at lightning speed, not waiting weeks for a new app build.
Inworld Runtime is engineered to tackle all three for developers. At its heart is a system called ‘Adaptive Graphs’, an engine room that lets an application grow from ten to ten million users without needing a complete architectural rebuild. It manages the flow of data between all the different AI services you might need, from language models to voice generation.
On top of that, it provides automated MLOps, acting like a guardian angel for your application. It automatically handles provider outages, manages performance spikes, and gives you a clear dashboard to see what’s going on under the bonnet.
Finally, its ‘Live Experiments’ feature is a dream for product teams, allowing them to test new ideas, prompts, or models with a single click, getting real-time feedback without ever needing to submit a new version to an app store.
“We scaled from prototype to one million users in 19 days with over 20x cost reduction,” commented Fai Nur, the CEO of WishRoll.
Streamlabs, meanwhile, used Runtime to build a streaming assistant with features that simply weren’t possible six months ago. The developers of Bible Chat slashed their voice generation costs by 85 percent while improving their features using Inworld’s platform.
Inworld Runtime’s pricing is based on usage, so developers can play around and experiment without huge upfront costs. It’s available with a huge choice of AI models and can be hooked into existing cloud accounts on Azure and Google Cloud, offering a practical path to get that next great idea into the world.
(Photo by Andrew George)
See also: React Native 0.81 brings platform consistency and faster compiles
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.
Explore other upcoming enterprise technology events and webinars powered by TechForge here.
