Meta AI on WhatsApp
Making AI feel native to messaging, for billions of people.
Senior Design Director
·
·
2023–2026
Shaped how AI appears in messaging on WhatsApp and shipped a broad set of AI features, from the first consumer launch through Incognito Chat.
Problem
When Meta set out to build a widely used AI assistant, WhatsApp had the largest audience and a high bar for privacy and craft: end-to-end encryption and a brand built on restraint. The opportunity was for WhatsApp to lead the company’s consumer AI rather than follow it. That came down to a few things: design interaction patterns native to messaging, strong enough that other Meta apps would adopt them; introduce AI in a way that fit the brand; and move at pace while protecting what makes WhatsApp trusted.
Approach
I operated at two altitudes: shaping product direction through CEO reviews (presenting strategy, defending tradeoffs, and translating feedback into shipped product) while building and running the design team. WhatsApp’s constraints of privacy, simplicity, and trust produced better product thinking, and the resulting patterns transferred to other apps.
Process
2023: Foundation
Created the mental model for how AI agents appear in messaging and presented the framework in CEO review; it helped guide AI product decisions across Meta’s apps. Scoped and shipped the initial consumer AI experience on WhatsApp, timed to Meta’s annual hardware moment.
2024: Surfacing AI
Defined the entry-points strategy for bringing AI into the messaging context; the pattern was later reflected across Meta’s other apps. Designed a voice interaction model that drove meaningfully deeper engagement than text and was subsequently adopted more broadly.
2025–2026: Private AI
Led the strategy for bringing AI into encrypted chat, a challenge specific to an encrypted app, shipping message summaries and writing help and culminating in Incognito Chat, the first fully private way to chat with AI.
Outcome
Patterns I designed (AI entry points, voice interaction, and Private Processing) were reflected across Meta’s other apps. Shipped a broad set of AI features spanning consumer AI, voice, multimodal, Private Processing, and AI Studio. Built and led the design team from the ground up through significant organizational change.
Meta AI chat
Lessons
Designing under tight constraints produced patterns that transferred well to less-constrained apps. Encryption, simplicity, and user trust forced sharper product decisions.
Recognition
Incognito Chat was covered by 200+ global press outlets at launch.
Press