Eugenia Kuyda’s latest AI product revisits an idea she first presented years ago: an assistant that helps with practical parts of daily life. Wabi is now available to all users, combining a messenger-style interface with task-focused threads, custom apps and conversations that can include other people.
From lifestyle concierge to personal assistant
Twelve years ago, Kuyda introduced IO at TechCrunch Disrupt. It was a conversational chatbot from her startup Luka, designed to help users book restaurants and vacations and handle other tasks. The product arrived before the deep learning advances behind today’s large language models.
Kuyda told TechCrunch that IO worked, though it was far less capable than current AI. Luka later became Replika, the AI companion company. After leaving her role as Replika’s CEO last year, Kuyda founded Wabi.
Wabi’s launch reflects how the original assistant idea has changed as the technology has developed. The product now enters a field that includes personal AI assistants such as Muse and Instinct, which aim to help with tasks including inbox management and finding flight deals.
Threads give tasks their own space
Wabi is organized as a messenger app where users can keep multiple AI threads for different tasks and longer-term goals. A health thread might include a weight-loss tracker, while a family thread could hold an app for finding children’s events nearby.
Users can create apps by asking for them or start new threads. Wabi can also generate a starting set of threads and apps when a person connects an email account. Kuyda described this as a shift from the company’s earlier approach, which emphasized making mini apps with prompts that people could share.
The team found, she said, that people value personal software over one-size-fits-all options, but may not want to build that software themselves. In Wabi’s current design, the assistant takes on more of the app creation, while users can request additions as their needs emerge.
That organization is also meant to address the limits of keeping everything in one long conversation. Kuyda argues that a single thread can become difficult to use as information accumulates. Separate threads give different goals and tasks a place of their own.
Shared chats connect the assistant to a group
Wabi’s threads can be shared, turning them into group chats where people and the assistant can coordinate. Kuyda sees this multiplayer structure as a key part of the product: instead of each person working with a separate AI, a group can use one shared conversation.
She showed TechCrunch a thread shared with her boyfriend. Its apps included a live music tracker, a daily relationship question, restaurant suggestions, events from their shared calendar and a list of books they were reading together, with recommendations for what to read next.
A family thread included a schedule tracker for Kuyda’s daughter, a finder for nearby kid-friendly events and a Japanese quiz for her daughter. These examples show how Wabi’s apps can sit inside the conversations where people already plan and coordinate.
People in a shared chat can choose what information to share, including email, calendar or bank account details. Kuyda said access is scoped to each individual chat, and participants can see when someone’s personal information is requested. The feature makes privacy settings part of how a group uses the assistant, rather than treating every thread as having the same access.
Privacy, purpose and an unsettled business model
Wabi is SOC 2 compliant, meaning its data security controls have been independently audited. Kuyda said the service runs entirely in the cloud and stores most user data in the U.S.
The product also carries forward part of the aim Kuyda associated with Replika: using AI to help people feel less lonely. In Wabi, she sees that potential in helping people spend time together offline, whether they are finding a concert to attend or choosing a book to read together.
Wabi is free for now, and the company has not settled on a business model. Kuyda said it is in talks to raise more funds. For the moment, the product’s central proposition is that an assistant can do more than answer individual requests: it can organize personal tools around ongoing goals and bring other people into the work.