Cal AI Founder Raises $10M for New Personal AI Startup Persona

The landscape of artificial intelligence is rapidly evolving, with young entrepreneurs often leading the charge into uncharted territory. Zach Yadegari, who made waves as a teenager by building and successfully selling the viral calorie-tracking app Cal AI to MyFitnessPal for an undisclosed sum after generating over $30 million in annual revenue, is back in the spotlight. After leaving his role at MyFitnessPal in June, Yadegari immediately set to work on his next ambitious venture. That vision has materialized as Persona, a personal AI assistant startup that has already managed to secure $10 million in funding during its initial round led by Vine Ventures, with participation from Z Fellows founder Cory Levy and Collective Global.

The core offering of Persona is a personal AI assistant that bears similarity to existing products on the market like Instinct or Meta’s Muse, but it stands out by incorporating a dedicated hardware component. The startup is preparing to release a $179 wearable band scheduled for availability in December, placing it in a similar category to the Bee wearable device that Amazon acquired. However, Yadegari points out distinct operational differences. Unlike some ambient competitors, Persona's band is not designed to continuously listen and record everything in the user's environment. Instead, the assistant activates through a physical button press or a deliberate flick of the wrist, giving the user complete control over when the AI is actively engaged.

Privacy and security represent major pillars of Persona's architecture. Yadegari emphasizes that the startup is engineered from the ground up to protect user data. The device itself does not store user information or run local AI models; rather, processing occurs securely within Persona’s cloud infrastructure. All user data is encrypted both in transit and at rest, and the company provides mechanisms for users to permanently delete their information. Furthermore, Persona promises that conversations will remain strictly confidential and will never be shared, traded, or sold to third-party advertisers or data brokers, addressing a growing consumer concern regarding digital privacy.

Security measures also extend to protecting the agent from external vulnerabilities and unauthorized access. The development team is actively implementing robust safeguards to prevent common AI agent pitfalls, such as phishing attacks and malicious prompt-injection attempts. Financial security is another critical focus area. Yadegari reassures users that Persona’s agent is architected so that it cannot access sensitive payment details like credit card numbers. Any financial transactions or purchases will strictly require explicit user approval and will be processed securely through external payment providers such as Stripe Link.

Despite a firm commitment not to sell user data to advertisers, Persona's preliminary business model does incorporate targeted advertising, specifically within the context of shopping research. When a user queries the assistant for product recommendations, the AI will factor in individual preferences while seamlessly integrating sponsored results. Yadegari explains that the recommendations remain unbiased because the underlying agent is intentionally designed not to know which specific items are sponsored placements. For instance, searching for office chairs will yield tailored suggestions matching user tastes, with some options being sponsored seamlessly.

As the hardware band approaches its December release date alongside a dedicated companion app, Persona has already launched as a free beta product. Users can currently interact with the AI assistant directly through iMessage text messages. This early phase has already attracted thousands of beta users and generated five-figure pre-order revenue for the wearable band. By blending accessible software interactions with a privacy-focused wearable device, Yadegari and his team are positioning Persona to carve out a unique space in the crowded consumer AI landscape.