AI startup Wabi is changing the way users interact with its technology. The company, which previously focused on allowing people to create applications through natural-language prompts, is now moving toward a more conversational AI experience that can also build custom interfaces when users need them.
The new product, called Wabi 2.0, arrives as AI agents become increasingly focused on completing tasks rather than simply answering questions. Wabi wants to combine the convenience of an AI assistant with the flexibility of traditional software.
From Prompt-Based App Creation to Personal AI Agents
Wabi was originally designed around the idea that anyone could describe an application they wanted and have AI create it.
With Wabi 2.0, that capability remains, but the starting point has changed. Instead of asking users to think about which application they need to build, Wabi positions its AI as a personal agent that can determine when a custom interface would make a task easier.
The company was founded by Eugenia Kuyda, who also founded AI companion company Replika. She describes the new experience as a personal agent capable of performing tasks and creating the interface required for the situation.
This puts Wabi closer to the growing category of AI agents that combine conversations, automation, and productivity tools.
Why Chat Alone May Not Be Enough
Wabi’s shift is based on a simple observation about conversational interfaces.
A chat window works well for asking questions or giving instructions, but long-running projects can become difficult to manage when everything is stored in one continuous conversation.
Important information can become buried, while frequently used functions may require users to repeatedly describe what they want.
Wabi’s new approach attempts to solve this by combining messaging with interactive software.
Instead of only responding with text, the agent can create an interface designed around the user’s current objective.
Software That Appears When You Need It
Imagine discussing fitness goals with an AI assistant. Rather than simply receiving suggestions in a conversation, the system could create a calorie tracker, workout log, or weight-monitoring tool that remains accessible within the relevant conversation.
The same concept could apply to family and household activities.
A conversation about parenting could produce a calendar for children’s activities, a language-learning tool, or a way to track nearby family-friendly events.
These tools do not necessarily have to exist as traditional standalone applications before the user needs them. The agent can generate the appropriate interface as part of the interaction.
This creates a hybrid model in which conversations become the starting point, while applications provide the structure needed for recurring tasks.
A Broader Shift in AI Interfaces
Wabi’s direction reflects a larger change taking place across the AI industry.
Companies are increasingly exploring ways to bring software directly into AI assistants instead of requiring users to move between separate applications.
OpenAI, for example, has been expanding its ChatGPT plugin architecture to support more interactive application experiences inside conversations.
Wabi is approaching a similar problem from another direction. Rather than simply placing existing applications inside an AI assistant, it wants the agent itself to create software experiences based on what the user is trying to accomplish.
The company believes the future could combine three elements in one environment: AI agents, personalized applications, and the people users interact with.
That could represent a meaningful change in how consumers think about software. Instead of downloading an application first and then figuring out how to use it, users may increasingly begin with a goal and allow AI to create or connect the tools required to achieve it.
Wabi 2.0 Starts With an Invite
The new Wabi 2.0 experience is currently being offered through an invite-only model, with access codes being distributed through social media.
Its evolution will be worth watching as AI companies compete to determine whether the future of personal software is built around traditional apps, conversational assistants, autonomous agents, or a combination of all three.
