The least glamorous object in the creator economy is the editing timeline. It is a gray trench of waveforms, cuts, caption tracks and tiny decisions. Every minute of finished video can demand an indecent number of minutes staring back at it. Grace Wang built her career close enough to that trench to notice the real shortage in online video. People were not running out of stories. They were running out of hours.
Wang is the co-founder and chief marketing officer of OpusClip, the software company that asks a long video to reveal its shorter, more portable selves. Feed the system a podcast, an interview, a webinar, a sports recording or a vlog, and it searches for moments that can stand on their own. It can reframe a shot for a vertical screen, add captions and help prepare the result for social feeds. The proposition sounds almost cheeky: keep the story, lose the repetitive labor.
Her route into this work was not the standard pilgrimage through film school or an editing bay. Wang studied finance at Peking University, also attended the University of San Diego, worked as an associate at StratOp Group, and then moved into user growth and go-to-market strategy at ByteDance. Those chapters share a concern that now sits at the center of OpusClip: how an idea finds an audience.
Before the software, there was the queue
The decisive education came closer to the mess. Before founding OpusClip in 2022, Wang and co-founders Young Zhao and Jay Wu led a 500-person social-media talent agency. At that scale, post-production stops looking like a creative afterthought and starts looking like industrial plumbing. Recordings accumulate. Platforms demand different shapes. A useful exchange at minute 42 remains invisible because nobody has time to find it, cut it and publish it before the conversation moves on.
Wang summarized the bind with plain impatience: “People just don’t have resources, or it wastes so much time to do this before.” The line lacks the usual perfume of a startup slogan. That is why it works. Editing is expensive even when the software is cheap, because attention is the invoice.
OpusClip was not the trio’s first answer. The team initially worked on an AI livestreaming product. The idea struggled to hold users. The clipping function inside it proved more useful than the larger machine around it, so the founders followed the part people wanted. This is the sort of pivot that looks beautifully inevitable only after it succeeds. At the time, it meant admitting the ornate plan was less valuable than one stubborn utility.
A launch held together by Discord
The early public version of OpusClip arrived through Discord, borrowing the distribution pattern that had helped Midjourney place new AI behavior inside a familiar community. The company reported 1,000 users in three days. Within weeks, the figure had climbed to 200,000 creators and $1 million in annualized revenue. Users were not merely tolerating the rough edges. Some asked how to pay for more downloads, better resolution or a shorter queue.
Fast adoption can be a noisy fact. The more revealing detail is what people wanted to buy: not spectacle, but relief. They wanted the queue to move. A creator who has already done the difficult work of recording a thoughtful hour does not necessarily dream of becoming an expert in aspect ratios. The product met a job that was painfully easy to describe.
Wang’s own early notes about short-form video were correspondingly practical. A clip needed a clear hook in the first three seconds, a concise body that rewarded attention, and a strong call to action. She worried about creators being told to produce at a pace that invited burnout. The marketer’s work, in her telling, was not to sprinkle virality over indifferent material. It was to give worthwhile material a structure the feed could recognize.
Long recording
Find moments
Reframe and caption
Publish and learn
The CMO at the listening post
At VidCon in 2023, Wang treated the OpusClip booth as part stage, part field station. The product had been chosen for the event’s Innovation Showcase. There were demos, creators and the delighted silliness of asking someone how much she would invest if handed an imaginary venture fund. A YouTuber Wang admired replied that she would put in all of it. Wang shared the moment as an ego boost, then quickly returned to what mattered: feedback and suggestions from creators and investors.
That instinct places her role somewhere between chief marketer, customer translator and talent scout. Former OpusClip executive Franz-Josef Schrepf recalled that Wang contacted him within hours of his previous company being acquired. He used the story to describe a company that watches for strong people and moves quickly. Marketing here is not a department that waits for a finished product. It is a sensor attached to the company.
The founders also entered AI Grant’s second batch, run by Nat Friedman and Daniel Gross. Wang’s public response looked forward rather than ceremonial: “What comes next? Let AI editor shoulders the weight, liberating human creativity. From 1 million to 100 million creators and businesses.” The grammar is less interesting than the scale. She was already describing editing not as a specialist’s workstation, but as infrastructure for anybody with something to say.
Millions of private editing rooms
By August 2024, OpusClip had announced $30 million in total funding, including a Series A led by Millennium New Horizons. In March 2025, it announced another $20 million from SoftBank Vision Fund 2 and others. The company said more than 10 million creators and businesses had used the product by then, making 172 million clips in 2024. Those clips had accumulated more than 57 billion views, according to OpusClip.
The customer list widened with the numbers. OpusClip named Univision, Billboard, iHeartMedia, Visa, LinkedIn, Jubilee and Impact Theory among its users. That creates a peculiar product challenge. A solo creator publishing a first video and a media company searching a vast archive do not look alike from the outside. Inside the editing room, however, each is trying to make existing work travel farther without multiplying the hours required.
The product expanded around that shared constraint. ClipAnything was designed to work beyond talking heads by reading visual, audio and sentiment cues. OpusSearch was built to retrieve moments from a catalog. Reframing, captions, B-roll and templates addressed the small frictions that turn one export into an afternoon. The company’s creator advisory board made the feedback loop more formal, placing working creators near product decisions.
There is an honest tension here. A system that promises “viral” clips is making a prediction about a public that remains spectacularly moody. Automation can find a clean hook; it cannot order millions of people to care. Product reviews reflect the bargain. Many users praise the time saved, while others want more control, steadier framing or fewer limits. The timeline may be tedious, but it is also where taste lives. OpusClip’s long-term test is whether it can remove drudgery without sanding away judgment.
The archive gets a second life
Wang’s distinctive contribution is to treat the archive as a growth problem. Her finance education supplied one vocabulary for underused assets; ByteDance supplied another for attention; the talent agency supplied the bruising reality of making enough material for modern distribution. OpusClip combines the three. A recording is no longer finished when it is uploaded. It is inventory, full of moments that may have missed their first appointment with an audience.
That idea changes the meaning of scale. The usual image is more cameras, more shoots, more people and more noise. Wang’s version begins with work already done. One conversation becomes several entries into the world. A lesson finds the student who would never watch an hour. A joke escapes the episode that hid it. The creator need not manufacture a new self for every platform; the existing story can simply wear more than one coat.
The aspiration is generous, though the work is exacting. To carry the weight, the software must understand which weight is dead and which is meaning. It must know that a pause can be awkward or electric, that the best reaction may happen before the sentence ends, and that a hook without a payoff is merely a door painted on a wall. Those are technical problems with editorial consequences.
Wang does not present herself as the lone genius in this story. Her public posts point outward - to creators, co-founders, teammates, investors and the audience itself. That fits the company she is building. Every useful clip is a collaboration among the person who spoke, the person who shaped the product, the machine that searched and the viewer who decided to stay.
The editing timeline will not disappear. People with taste will still argue over frames, rhythm and the exact half-second when a glance becomes funny. But fewer stories may die there from simple exhaustion. Wang’s wager is that the machine can handle the first search, the first cut and the repetitive resizing, leaving people to make the decisions that feel like decisions. It is not the end of editing. It is a shorter walk to the interesting part.