In just a month, a new production team leader switched the team to complete AI automation and halved art production time.
Process automation and content generation are two primary cases for generative AI in business, according to the Eugina Jordan Survey on Gen AI for Business for 2025.
While there is a debate about finding a balance between human creativity and AI automation, the technology can enhance the creative workflow. Some game studios reported that they reduced the time required to produce game assets and complete animations by 40%.
I joined the art team as a new team leader in a game development studio and jumped into a risky experiment. At the social R&D team, we fully replaced manual drawing with AI and cut art production time by 50% in just 4 weeks. I had to bring a new approach to the mature team, rethink our processes, discover solutions, test, and implement them in our workflow. Here’s how we pulled it off.
The Challenge: Change a Running System
I have a background in leading creative teams for more than six years, though I joined Nexters during a major restructuring. It was a curious time: on the one hand, I could rethink and redo our production pipeline. On the other, the processes were already well-established. We had to find a middle ground.
My team is made up of seven strong professionals, artists, and animators, and we have also brought in an AI specialist to help with unexpected bugs and script selection. They produce high-quality ad content, very polished, literally mini-animated films. However, the market changes and demands something new.
According to Gitnux, 70% of studios report an increase in creative output after implementing AI into their workflows. Our goal was to cut the time from idea to UA testing for video creatives from 4 weeks to 2. How could we replace all the manual art with AI without dropping quality? We agreed from the start that even if we sped up processes by 20%, we’d win.
3 Workflow Changes That Made a Difference
I spent three months as a team leader, during which I studied the existing processes. Then we received a business request to speed up production, and I initiated the changes.
The first and most obvious was to parallelize processes. We used an approach similar to creative pairing. The writer (director) and the artist worked together from the very first brainstorming session. The artist made sketches at the idea stage so that the concept already had a visual form from the very beginning.
We worked in cycles. For example, while one set of tasks was being done, another specialist would start the next cycle. Now one artist can complete two cycles of tasks in a week.
We used to go through all the material in one cycle and make edits immediately, aiming for a final “perfect” version. Now, if something doesn’t work from a hypothesis perspective, we improve it in the next iteration.
Previously, production was linear: first the idea, next the art, then the animation. By combining some of these stages, we saved about 25% of the time.
Second, we implemented overlapping animation and art stages. For example:
- Monday: full-day ideation and scriptwriting
- Tuesday: team briefing + artist starts sketching during the briefing
- Wednesday: the artist and director continue to refine, and some assets are ready to be transferred to animation
This approach helped us:
- Spot implementation challenges early
- Simplify complex elements
- Finalize the storyline early and avoid late-stage rewrites
Third, we reused existing assets, as the team had a rich art archive. Directors adapted storylines, and our artists knew the database by heart.
AI Tools in Action
Regarding the AI tools, we used ChatGPT for prompt generation. Sometimes it showed great results even on the first try, which is rare considering the specific art style. Stable Diffusion and Krea.ai worked well to generate characters, environments, or objects in a specific style. And with Moho, we sped up some animation tasks.
Within just two weeks, we saw the first results: the overall production cycle was cut from four weeks to two, and static material creation now takes only 2.5 to 3 days, twice as fast as before. We also reduced animation time from 10 to approximately 7 working days by reusing assets and simplifying storylines. What’s more, performance did not drop. The slight decrease in quality did not hurt the results.
Overcoming Resistance to Change in a Team
It’s challenging for a new team leader to sell changes to a mature team.
First, I showed up with excitement in my eyes and said, “Look, I ran the numbers. We can produce more, deliver, test, and find our direction faster.” I saw the team sitting there like, “Okay… why do we need this?”
I realized: my teammates had worked in the company for years and truly wanted it to succeed. However, they didn’t understand the connection between a short-term goal and the bigger picture.
I had to explain that we weren’t going to work more, but smarter. We see the team as a long-term investment, and we don’t want to burn them out in two months. I took routine tasks off our strongest specialists so they had more time to learn and experiment with AI. In our team, senior artists and experts mentored the others, reviewed work, tracked progress, and spent 30% of their time on this type of work.
Another concern was AI itself. The team was skeptical about it at first. Imagine the artists, who were used to working to very high standards, creating ads as short animated films, now being asked to do everything with AI, which is unpredictable and unreliable under tight deadlines.
To overcome such concerns, I believe it is crucial to provide AI learning programs to the team. Nearly half of the employees surveyed by McKinsey stated they want more formal training and see it as the most effective way to drive AI adoption.
How to Promote New Ideas to a Team: What Worked
It’s important to be on the same page with the whole team and stay informed about how things are going. Here I explain negotiation tips I used with my team:
- Clearly explain the goal. For performers, the bigger goals aren’t always obvious; you need to break them down and translate them into actionable bits.
- Be honest. In my case, it was important to openly address the fears employees were probably keeping to themselves. Show them you see it, that you’re not just their boss saying, “Let’s do it,” without a plan.
- Collect feedback. If things had gone completely off the rails, I would’ve needed to apply different tools. There were tricky moments, for example, when our AI tool turned a human character into a cyclops. We created a small chat group where we shared all the weird stuff AI was generating, and the AI consultant helped us quickly address unexpected AI outcomes.
- Highlight each team member’s personal interest in the work. Some people had fears like, “If I don’t learn to use this, I’ll be fired.” And I’d say, “If you do learn, you’ll become a super valuable team member”. For me, those one-on-one meetings are all about finding each person’s passion and figuring out how it can align with the company’s goals.
- Wrap things up and thank people. It’s essential to make the experience feel rewarding and repeatable.
What’s Next for Our Team, or How We Help People Grow
It’s clear that AI will continue to develop, and so we have to test and implement the most effective practices for working with this technology.
Teams using AI were three times more likely to create top-performing solutions than traditional ones, states recent research by Harvard Business School and Wharton.
We work within R&D, so we strive to run as many tests as possible to find solutions that work. Our focus is not on perfection but on achieving a level of quality good enough to test ideas. This AI experiment was successful, and during monthly reviews, I see that our creatives are performing well. We actually have things to celebrate.
The team learned many things. What’s more, two months later, I still see the team using AI and scripts. They’re even sharing knowledge with other teams. I want this kind of work to be something we’re all proud of.
What’s next? How I plan team development depends a lot on the overall marketing strategy. My team is small, so I’m focused on stabilizing things during changes and making sure everyone has area to grow.
Let me explain with an example: we have a wonderful artist who gave great feedback on art direction. She worked with AI tools, realized the possibilities of the technology, and afterward became more interested in processes in general. Now we’re creating a development track for her focused on process improvement.
Currently, our team development plan is all about identifying and strengthening those interest points for each person on the team.
