AI Tools Artist Spotlight: an Interview with Kelly Boesch by Jeff Foster

Support ProVideo Coalition

Shop with

Filmtools

Filmmakers go-to destination for pre-production, production & post production equipment!

Shop Now

An interview with one of AI’s creative visual pioneers.

Jeff Foster

January 5, 2026

Comment

AI Tools Artist Spotlight: Kelly Boesch

Kelly Boesch is an AI visual innovator blending two decades of design expertise from IMAX and abstract painting with cutting-edge technology.

A pioneer in developing a new visual language, her visually arresting content has captivated over 2M followers and generates 60M+ views per month across platforms.

She sees AI as a powerful collaborator, pushing creative boundaries at the intersection of human artistry and machine learning.

We are truly fortunate to get a little insight with Kelly regarding her approach and process in developing one of AI’s most innovative and growing visual and aural escapism artform video channels.

An Interview with Kelly:

Q: Thank you for agreeing to this interview, Kelly. Can you please tell us a bit about yourself in general?

I am a visual artist, graphic designer and video editor based in Los Angeles. My passion has always been rooted in the arts. I have a background as an abstract painter using acrylics and mixed media, which heavily influences how I see color and composition in the digital space.

I’m a bit of a creative sponge—whether it’s painting, jewelry design, or now AI, I’m always looking for new ways to tell stories visually.

I am an animal lover and try to be a caring and kind person who still relies on empathy, even though I keep hearing it’s out of style.

I am also a music lover. I have always been drawn to music from the 90’s, downtempo, acid jazz. Cool music with a groove. I think that’s what helped when I started using Suno to make music for my videos. I jumped right in and was able to create music that complimented my videos very well.

Q: What has your career path been like?

My career has been a mix of high-end entertainment film, fashion and corporate stability. I worked for several years for a fashion company running a design studio and running a CAD design computer. I spent 17 years at IMAX in film production and marketing, which really laid the foundation for my visual approach. I learned how to think “cinematic” and big-picture there.

I have always been a creative person. I studied fine arts in college and have always had some type of creative side to every job I have worked at. For IMAX I used to create the development decks that we used to raise sponsorship money. This was way before AI so I had to get really creative.

I also have freelanced off and on for many years. My work as a freelance graphic designer consisted of making music videos for music artists, designing album covers, all sorts of things. The moment I saw AI though, 3 years ago, I was instantly hooked and have been completely focused on it as my chosen art form.

Q: How long have you been producing with Gen AI tools? How did you get started?

I’ve been honing my skills in AI art for about three years now. I started, like many of us, with static image generation. I was fascinated by the ability to visualize the surreal concepts in my head instantly. As soon as the video tools (like RunwayML and Pika) started becoming viable, I jumped in. I realized I didn’t just want to make pictures; I wanted to make music videos. It’s been a non-stop sprint of learning ever since. The ability to make your own video content just blows my mind. I used to have to go out and shoot things or buy stock videos. Now I can create anything I want sitting on my couch. It’s mind blowing.

Q: Do you have a specific approach to creating an animation video?

I almost always start with a strong visual concept or a specific set of characters. Because of my painting background, I’m very particular about composition and color palettes.

I usually develop the “hero” image first—the character or the setting—using Midjourney. Once I have a static image that evokes the right feeling, I let that dictate the motion.

I don’t usually use text to video; I prefer Image-to-Video workflows because they give me the control I need to maintain my specific artistic style. I have been playing around a bit with Nano Banana but for the most part I stick with Midjourney. The creativity in that app is astounding.

Lately I have been animating mostly with VEO3. That is one app I have to pay for. I am creative partners with RunwayML, Higgsfield, Hailuo and Pika so I do switch around a lot and post and tag to keep up my partnerships. VEO3 is so cool though as it also gives you sound and follows prompting so good.

You can be so specific with the prompts and it follows right along. Also, I tend to stick to the creative/artistic side of AI. Every now and then I play around with realism but it doesn’t really interest me.

I have no interest in making ai films. I find it tedious and boring. I see what some other artists do in this area and am blown away. I know how hard it is. My talents lie in these more music video type formats.

Q: You mentioned in one of your posts on LinkedIn that you typically create the videos first and then the music. The music always fits the videos and characters so well – especially the dance moves! How do you approach the syncing?

It sounds backward to some, but the visual rhythm comes first for me. I generate the clips to get the movement—the swagger of a character or the flow of a dress. Once I see how the character “moves,” I can usually tell what style of music I want to use.

If I am going to use lyrics in the song I sit and write out the story line for the lyrics, or theme or whatever I want the lyrics to tell. I am not a writer but I have good ideas. I sometimes think of a good hook for a song, or idea. I write it all down in train of thought and then ask AI to put it into lyric format and then I can rewrite the song.

AI tends to use the same lyric words over and over and they always sound alike. Best to take the time to write out the lyrics the best you can to get a better song. I then use Suno to generate a track that matches that specific tempo and vibe. I love to mix styles. I go crazy. I try anything. Dance, Jazz, Polka, whatever. I just like to get weird with it. But sometimes I write out exactly what I have in mind and have AI write a Suno prompt for me based on the idea. That will include the tempo, instruments, style, etc. It’s always different.

Every now and then I get an idea for a song during the day and I make the song first and then make a video to go with the song. For some reason I find this much harder. Not sure why. I always edit the videos and music together though. I lay down the track and choose what videos go where.

The “sync” happens in the edit; I use Adobe Premiere to cut the video tightly to the beat, which makes it feel like the character is reacting to the music, even though the music came second. Or if it’s a dance video I try to match the movement to the music in sections. It’s a process and mostly done by feeling. Which I think people pick up on.

I get messages every day from people telling me that my video really made them “feel” something. I LOVE that. Makes my day. I get the most incredible, personal messages from people. Reminds me that there is still kindness in the world. I try to read and respond to as many as I can. I post on 6 apps though and get thousands of comments and messages daily. And if I get behind then it’s impossible. I have a full time job still so I can’t do any of this until after work so things have been a bit tricky lately. I still need time to make my daily creation.

Q: Can you walk us through a typical workflow and lead us through the breakdown of a project?

My workflow is a hybrid of AI and traditional post-production.

Ideation: I create the base assets in Midjourney. This is where I establish the “look.” I either have a specific idea or I play around using old images of mine, strange images of insects or other things and just play until something sticks and I get an idea. My videos always have a theme or a look.

Prep: I often take those images into Photoshop (using Generative Fill) to expand borders or clean up details so they are ready for animation. I upscale using Topaz Gigapixel if needed but they don’t always need it.

Animation: I import those images into video generators. I use a mix depending on the need—VEO3, Higgsfield, Midjourney, Hailuo or RunwayML or Pika. Pika is really great for keyframe videos, the ones that morph. I do a lot with that style. Midjourney is pretty good too. First frame/last frame is another name for it. I can usually tell how each app will animate a certain type of image so I base my animation choice on that.

Music: I generate the soundtrack using Suno.

Assembly: I bring everything into Adobe Premiere Pro. This is crucial—you can’t just let the AI do the work. The editing, color correction, and sound mixing are where the “human touch” turns a clip into a story. Sometimes I drop the entire completed video into CapCut and add a slight final filter to the whole video. It really gives it a nice, complete look

Q: Tell us what your plans are next.

I am focused on pushing the boundaries of what this “new visual language” can be. I’m currently a creative partner/ambassador for several platform, so I plan to keep testing their limits. I have been getting the most amazing offers to speak all around the world. I may go to Hong Kong in March, Cannes in April. And in London I will be the featured artist on the digital billboard/building on Oxford street with W1 Curates. They featured me in August of this year at two of the biggest clubs in Ibiza.

I just had an art show at a Museum in Mount Vernon, Il, The Cedarhurst Center for the Arts. They are definitely trail blazers. They received a lot of flack about doing an ai show but it actually went over very well. I have also been featured in several magazines this year.

My goal is to continue proving that AI is a collaborator that augments human creativity, rather than replacing it. I really want to focus on my own work and less on other projects. I feel like I have been taken away from what I love to do so much lately. Creating is what brings me the most joy.

I want to continue to build my YouTube channel. I actually am making money there now. A lot of people watch my videos on their TV, maybe when having parties or just hanging at home. I have made a lot of playlists so people can choose their vibe. YouTube works best on engagement.

Kelley’s Links: