A startup called Empromptu Inc. is looking to deliver where all other artificial intelligence-based application building tools have failed. It’s launching what it says is the first platform of its kind to deliver complete, business-ready AI applications that can be relied upon to work in production and integrate smoothly with existing technology systems.

Empromptu reckons that the vast majority of AI-based no-code and low-code application builders fail to deliver the goods, pointing to a study by the Massachusetts Institute of Technology that shows how 95% of AI pilot projects ultimately end up being thrown onto the scrap heap.

The reasons for these failures are extensive, but in many cases it’s because companies don’t have the resources needed to hire dedicated and experienced large language model engineers to ensure everything works as it should. Moreover, most platforms rely on “generic” LLMs rather than models that have been fine-tuned for the application in mind, resulting in fragile tools that struggle to perform in production environments. This means that most new applications built using AI or “vibe coding” tools can only achieve accuracy of around 60% to 70%, which isn’t reliable enough for most enterprise scenarios.

Empromptu says its AI app builder does much better than this because it’s the first such platform that’s built on a complete AI stack, with LLMOps tooling, retrieval-augmented generation or RAG, automated AI response optimization and other essential features.

Enhanced accuracy

Just like many of the AI app building tools it’s trying to replace, Empromptu is aimed at regular business employees rather than developers, offering a no-code interface that allows them to describe the application they want to build using natural language.

From the user’s description, Empromptu will then set about building a complete, production-ready AI application based on a full stack that includes the front end, back end and AI logic, along with a user interface that integrates with the user’s existing workflow. With built-in automatic AI response optimization, Empromptu says, its generated apps achieve 98% accuracy in response to user’s prompts, thanks in part to its use of RAG, which makes it possible for them to tap into secure corporate data to enhance its responses.

Empromptu said the secret to its accuracy is its proprietary optimization technology, which was developed by its co-founder Dr. Sean Robinson while working at the National Aeronautics and Space Administration.

“There’s no shortage of tools that leverage AI to build an application, but building an ‘AI’ application is different,” said Empromptu co-founder and Chief Executive Shanea Leven. “Empromptu cuts through the vibe coding hype to make reliable AI applications a reality.”

From prompt to production-grade app

Using Empromptu, anyone can build an AI-powered app within just a few minutes. To begin, they just describe the kind of application they want to build, such as “I want to build a customer support assistant that understands our product documentation and provides accurate troubleshooting steps.”

The AI builder will then ask a number of questions to clarify exactly what the app is supposed to do. For instance, it may ask about the data sources the app should use, the way in which users will interact with it, what constitutes a successful response, and how the user intends to deploy the app. Armed with those responses, it will then generate a comprehensive blueprint of the app, based on its data requirements, AI tasks, integration specifications and optimization criteria, and then hand it off to a team of AI agents that actually build the app. Once done, users can then tweak the app through additional prompts to get it just the way they imagined.

Through this simple process, Empromptu says, its platform can be used to create extremely powerful AI applications for almost any use case. For instance, the educational organization Practicing Musician SPC used it to create an AI music application that supports students, educators, artists and parents by combining traditional teaching techniques with modern learning methods. It developed the app based on a single natural language prompt that outlined its intended purpose, before refining it with additional prompts.

Rory Creigh, Practicing Musician’s director of AI and education, said the end result was a complex educational app in a highly specialized domain. “Every time, we have seen 98% accuracy or better,” he said. “That reliability lets us serve the community with intuitive, research-based learning, expanding our impact and making music education more accessible.”

Image: SiliconANGLE/Microsoft Designer