Helena Zhang Helped Bring Creative AI Out of the Lab and Into Everyday Workflows

How an early OpenArt leader turned complex image and video tools into usable creative systems for millions

Creative AI no longer lives only in research demos, invite-only tools, or experimental corners of the internet. It is becoming part of how stories are made, products are tested, campaigns are built, and visual ideas move from a rough thought to a finished piece of content. Few people have worked closer to that shift than Helena Zhang, an entrepreneur, AI product engineer, and creative AI expert who helped make advanced generation tools easier for everyday creators to understand and use.

The change has been fast. Image generation was once the headline. Then came sharper editing tools, stronger video models, more realistic voice generation, and creative systems that let users move across formats. A creator can now generate a character sheet, turn still images into key frames, use those key frames to guide video, and add voice or sound to build a longer story. The workflow is no longer one prompt and one image. It is a chain of decisions.

For Zhang, that chain is where the real creative work happens.“People sometimes talk about AI as if the creative act ends when the model gives you an output,” Zhang says. “I see it very differently. The strongest work comes when a person keeps directing, shaping, and refining what the tool produces.”

That belief has guided much of her work. Zhang joined OpenArt as its second hire and head of growth, helping the company grow from zero to 60M ARR. OpenArt became known as a major platform for AI image and video generation, partnering with models including Seedream, Nano Banana 2, GPT Image 2, Seedance 2.0, Google Veo 3, and Kling. The company’s work placed Zhang close to some of the most visible changes in generative AI, but her contribution was not limited to bringing models onto a platform. She focused on helping people use them with more confidence and control.

That distinction matters in a field that can easily become technical for technicality’s sake. Many creative AI tools are powerful, but power alone does not help a user make better work. A small-business owner, filmmaker, designer, founder, or student needs to know what the tool can do, how to direct it, and how to recover when the first result misses the mark. Zhang’s work at OpenArt centered on closing that gap.

“The question was not only whether AI could generate something impressive,” Zhang says. “The more important question was whether a person could guide it toward what they actually wanted.”

One of the clearest examples came through OpenArt’s YouTube channel, which Zhang grew from zero to 200K subscribers in a year. The channel did more than promote features. It taught creative workflows at a time when users were trying to understand what was possible. Zhang created tutorials on training Flux Dev LoRAs to capture a specific character or style, including a video that showed how a model could learn to generate images featuring her dog. She also made tutorials on AI image editing and inpainting, showing how a creator could highlight part of an image and ask AI to replace or change that area.

Another popular tutorial focused on creating consistent characters with Kling 1.6 Images to Video without training a LoRA. That kind of workflow addressed one of the central frustrations in AI image and video work: maintaining character identity across scenes. For creators trying to tell stories, consistency is not a small issue. A character cannot look like one person in the opening frame and someone else a few seconds later. Creative control is what turns a model from a novelty into a production tool.

Zhang’s work made that point visible.

“A creator should not feel like they are waiting for a lucky result,” she says. “They should feel like they are directing the work. AI can be an amplifier, but people still need the ability to steer it.”

That philosophy runs against the easiest misconception about generative AI. Many people assume that progress means removing the human from the process. Zhang argues that the opposite is happening. As image, video, voice, and sound become easier to generate, human taste becomes more important, not less. AI can produce content that looks realistic because it has learned from massive datasets. What it does not automatically understand is why a story matters, what emotion the creator wants to evoke, or what an audience is supposed to feel.

In Zhang’s view, the human role is shifting rather than disappearing. Creators are becoming directors of increasingly capable systems. They need judgment, taste, patience, and the ability to recognize when an output is merely polished versus when it is meaningful.“

AI raises the floor,” Zhang says. “It makes it easier to get something that looks finished. But making something great still takes care, intent, and taste.”

Her authority in the field comes from working across product development, growth, education, and user behavior. At OpenArt, she helped shape features around actual creative needs. At Fish Audio, where she was a cofounder, she worked in another crucial part of the creative AI stack: voice. Fish Audio launched open-weight AI text-to-speech models supporting more than 80 languages, and users created more than 5 million voice clones. That experience expanded Zhang’s view of creative tools beyond images and video into the sound and personality of generated content.

Her work has also placed her in broader industry conversations. Zhang judged at the 2025 MIT AI Filmmaking Hackathon, was invited to speak at the 2025 AI User Conference, and was invited to speak at SendPulse and AiMe Academy’s AI Marketing Day 2025. She was also ranked the #65 most influential user on Product Hunt in 2025 and had four Top 5 Product of the Day launches.

Still, Zhang’s view of the industry remains grounded in users, not awards. She often points out that creative AI will only stay democratic if education keeps pace with the tools themselves. The technology is changing quickly. Without clear teaching, the most powerful workflows can become accessible only to insiders who already know where to look.

That is not the future Zhang wants.

She believes the next phase of creative AI should help more people express ideas with greater precision. That requires tools that are powerful enough for ambitious work, but clear enough for people outside the technical core of the industry. It also requires builders who listen closely to how people actually create.

“People around the world are using these tools in ways that product teams may not expect,” Zhang says. “Some of the best workflows come from listening to users and understanding the problem they are trying to solve.”

That listening now informs Zhang’s next chapter. After working on creative AI tools at OpenArt and Fish Audio, she is focused on bringing a more authentic voice to AI. Her work is moving toward systems that understand user intent, psychology, word choice, and longer-term creative goals, rather than simply producing content that resembles a training set.

The larger vision is not AI as a replacement for imagination. Zhang describes the better version as a friend, a creative partner, and a coworker, something that helps people make more of what they already carry.

The creative AI boom has produced plenty of tools that can generate. Zhang’s work points toward something more useful: tools that can be directed, understood, and shaped by the people using them. That may be the difference between a machine that makes content and a system that helps people create.