Hi there, this is Bin Liu. I’m the former co-founder and CEO of Alisa. Today, Alisa is officially part of HeyGen, and I’ve joined the team as the VP of Product Engineering. Together with Joshua and the HeyGen team, we’ll be building the next generation of the creative operating system.
How we got here
If Alisa is new to you, here’s a quick intro.
For the past 2.5 years, my team and I focused on one question: how can AI help everyday people turn ideas into finished work? Not just Q&A or asset drafts, but publishable work where creativity is the only limit. That question became our compass. We staked our mission on agent-driven content from day one and proved it by shipping, learning, and refining. We built our own intelligence stack required to achieve our mission, moved from early agentic workflows to interactive real-time agents, and ultimately to Alisa, the world’s first multimedia content agent.
Our roots shaped this journey. Most of us came from Pinterest, where we spent nearly a decade building products and systems that made hundreds of millions of people say “I’m inspired!”. As Head of Content & Creators, I discovered that many creators and entrepreneurs on Pinterest had brilliant ideas and brands but faced challenges in executing their creative vision. Many of them struggled to keep up with the fast-evolving content trends toward short-form videos. That gap became our mission and the spark for Alisa.
Since our Alisa beta launch this past April, we’ve listened to creators, entrepreneurs, educators, and marketers who are managing tight deadlines, numerous creative directions, and rising expectations for quality and engagement. Their struggle made one thing clear: the world does not need another visual toolkit. People need a partner that works like a teammate, understands context, adapts to your style, and helps you create ready-to-publish videos at scale.
To the people who trusted Alisa in our earliest days, thank you. You showed us what an intelligent, creative partner must be.
Why HeyGen + Alisa
Today, HeyGen launched Video Agent in public beta. Video Agent is your creative operating system. From a single prompt, it creates a complete, publishable video for all your needs. Script, casting, visuals, voice, pacing, captions, edit, and style. End-to-end. Unlike generators that output clips or scenes, Video Agent does what a creative team would in one intelligent workflow, right in your browser.
Video Agent is similar to what we were building at Alisa. Joshua and I have traded notes for years on product, company building, and the future of AI agents. After we launched Alisa, a single, long conversation made the path forward clear. And on that path, two things excite me the most.
First, Alisa + HeyGen hyper-realistic avatars
Think of Video Agent as the director and HeyGen’s avatars as the on-camera talent. Pair our agentic brain with their lifelike actors (including Avatar IV and Digital Twin), and you get performances that carry nuance: micro-gestures, eye lines, and natural pacing. It’s the difference between stock footage and a real spokesperson who never tires, never drifts off brand, and shows up perfectly every time.
Second, agentic intelligence on top of HeyGen’s creative suite
Adding Video Agent to the creative tools is like equipping your content filming studio with a ready-to-roll filming crew that listens to your creative directives. You provide a prompt, and Video Agent writes the script, edits the cut, translates the lines, adds captions, and iterates on your feedback in the same session. It also serves as a GPS for creative work: not just a map but turn-by-turn guidance to “publish.”
Together we’re building an AI agent that feels less like software and more like a creative operating system: thinking with you, drafting with you, and shipping with you. Not tomorrow. Now.
What we’ll be focused on
I envision a world where anyone can go from idea to publishing within the same day. Starting with just a prompt, Video Agent builds out a plan, picks the on-brand avatar and voice, pulls approved assets, sets pacing, adds captions, and outputs channel-ready cuts (1:1, 9:16, 16:9). Then it applies your brand kit, runs basic compliance, and translates for priority markets. You iterate with Video Agent for a few edits, hit publish, and push to LinkedIn, YouTube, and your CMS from the same screen. Work that took weeks now ships instantly.
Our focus is simple: build an AI agent that’s not just another generator. It’s a teammate that understands context, adapts to your style, and gets you to “done.”
If you have ever stared at a blank page with a great idea and not enough time, or if you’ve ever been told “this video is not a priority,” this is for you.
We're inviting you to co-create with us. Bring your use cases, your wish list, your toughest constraints, your insights, and your ideas! Together, we will make Video Agent your everyday creative operating system.
