We build mobile and web apps that make AI agents and generative AI usable for everyday consumers — not just engineers or power users. Three products live across iOS and the web today, with more in development.
A Delaware C-Corporation building consumer AI applications. We design, build, and operate our own mobile and web products end-to-end.
We turn frontier AI capabilities — agents, multi-agent reasoning, and generative video — into simple, mobile-first consumer apps.
Today's AI tools require terminals, prompts, API keys, and technical setup. A tiny slice of power users gets real value while the next billion people are locked out. We build the consumer layer that removes all of that.
Everyday consumers, creators, founders, and small teams who want AI to do real work — without learning new tools or writing prompts.
Active company. Three products live across the App Store and the web; one in development. Full-time team of builders with backgrounds at Meta, Google, and a prior consumer app that reached #1 on the US App Store.
AI today still lives behind a command line. Chat windows, terminals, copy-pasted prompts — a tiny slice of Silicon Valley calls this revolutionary, while the rest of the world is barely paying attention.
Every previous platform shift unlocked its real value the moment a great consumer interface arrived. DOS to Windows. Servers to mobile. Web pages to feeds. The same is about to happen to agents.
We build the products that bring AI agents to people who don't know how to write a prompt — and don't need to. Beautiful, opinionated, mobile-first.
In practice, that means we build apps — not frameworks, not infrastructure, not B2B tooling. Every product Leverage AI ships is a consumer-facing application that someone can download, sign into, and use today.
An AI-native workspace that gives agents real context, execution, and continuity across every machine you work on.
Agents today start from zero in every conversation. They lose context the moment a tab closes, can't reach the files and terminals where real work lives, and run in isolated silos that don't follow you between devices. Locus is the workspace platform that fixes this — a unified interface for files, machines, terminals, and browsers that any agent runtime can operate inside.
Folders, project state, terminal history, and prior sessions are exposed to every agent as a first-class workspace — not pasted into a chat box. Agents know what they're working on the moment they start.
A unified runtime spanning your local computer, shared workstations, and on-demand cloud machines. Terminals, browsers, and processes are managed by Locus and reachable from any device.
Persistent sessions survive closed tabs, switched devices, and handoffs between agents. Open a session on the web, continue from desktop, finish on mobile — work doesn't reset.
Four bets on what consumer AI looks like when the chat box disappears.
An AI-native workspace for agents. Locus understands your files, devices, and work history, so agents can act with the right context. Live on iOS and the web, with files, context, and workspaces syncing across mobile, web, and PC.
Visit locus.cool LiveMultiple AI agents in one room — debating, verifying, and cross-checking each other. Use it to pressure-test important decisions or research questions: agents argue, cite evidence, and corroborate before giving you an answer.
Open in App Store LiveRecreate any viral short-form video in seconds. CloneViral turns a single reference clip into a publish-ready video — no editing timeline, no prompts, no expertise required.
Try cloneviral.videoA general-purpose mobile agent that can act across your apps. Less chat, more outcome. Currently in development.
Coming soonA short walkthrough of the AI-native workspace — files, agents, and context across devices.
Previously founded LiveIn (2021–2022), a consumer social app that reached #1 on the US App Store and held the position for nearly two months, growing to 2M+ daily active users with virtually zero paid marketing. Before that, technical lead at Meta (2015–2019) on the Video Ads and Core User Growth teams, where his work contributed 20M+ DAUs to Facebook. MS in Computer Science from NYU; Stanford GSB Ignite Program.
linkedin.com/in/louis-xu-0a5ba545Previously a software engineer at Google (2017–2022), building production infrastructure. MS in Information Networking from CMU; BS in Computer Software Engineering from Wuhan University.
linkedin.com/in/wenjiet