In 1994, Rise of the Robots promised to be the future of gaming: cutting-edge 3D visuals, groundbreaking AI, and a great soundtrack. And then people played it.
What they got was clunky animations, lifeless characters, and duller-than-dull gameplay. Critics roasted it, players abandoned it, and Rise of the Robots went down in history as proof that cool tech means nothing if the experience sucks.
Fast-forward to today, and the stakes for visuals are even higher — especially in mobile gaming. Players demand breathtaking art and immersive worlds. Anything less is a speedrun straight to the app store graveyard.
The best game developers know this. Without strong visuals, it’s nearly impossible to attract new players, keep them engaged, or stand out in a crowded market. That’s why developers are always looking for ways to level up their designs without sacrificing quality, brand integrity, or valuable development time.
Layer.ai gets this. They’re revolutionizing game asset creation by blending professional game design with AI, giving developers the tools to create high-quality, custom assets in their unique brand style.
Their goal is simple: to work alongside game developers, helping them spend more time doing the work they love, creating art, without ever replacing the creative process.
How AI is Empowering Mobile Game Developers
Layer.ai has a bold mission: to make high-quality game art accessible to every developer. As VP of Engineering Joona Rahko puts it, "As a game developer, you have unlimited creativity, but often limited resources. We want to give every mobile gaming developer access to the best possible art, no matter their studio size or budget."
What makes Layer.ai unique is its ability to generate production-ready assets in a developer's custom brand style. Whether it’s characters, environments, UI elements, or marketing assets the platform ensures visual consistency without sacrificing creativity.
This is possible by having users “teach” the tooling by uploading existing artwork. Once input, Layer’s platform trains an AI model to match the art style, allowing teams to generate anything from new characters to game UI elements in a fraction of the time it would take to create them manually.
Transforming Game Design: Stories from Developers
Layer is actively reshaping how studios approach game development. Here are a few examples of how Layer’s customers are using the platform to level up their production.
Ace Games: Saving 120 hours a month
Ace Games, the team behind Fiona’s Farm, needed a new way to scale their in-house production pipelines for everything from character art to new level assets under high player demand.
By training Layer’s most popular AI mode, FLUX, on their existing game assets, they boosted productivity by 80%, saving over 120 hours monthly. This allowed them to quickly generate and refine assets for seasonal and event-based content, maintain visual consistency, and underburden their production team.
Tactical Strike: Saving 2500+ hours of production time
Lokum Games faced delays and inconsistent styles when outsourcing character illustrations for Tactical Strike. With Layer, they created 40 unique characters in under 30 days — saving hundreds of hours.
By training Layer’s AI on their existing artwork, Lokum saved 2500+ in production team bringing the power of creation in-house, allowing them to focus on refining other aspects of the game while maintaining visual consistency.
LBC Studios: 8x Boost in Productivity
LBC Studios, a medium size game studio, used Layer to create high-quality, on-brand assets for Brewtopia. What normally took 320+ hours for a character illustration was completed in just 40 hours with Layer’s AI.
The studio produced 10 characters, vehicles, and props in one workweek — an 8x productivity boost. The ability to export assets to Photoshop for final tweaks helped LBC maintain quality while accelerating the process and reducing costs.
Built for Trust
Beyond Layer’s commitment to innovation in mobile game design, they have a steadfast, ethical foundation that depends on data security and privacy.
Layer is the world’s first SOC 2-compliant gaming AI provider, a certification that ensures the highest standards of data processing and storage.
With a strong emphasis on privacy and compliance, Layer ensures that its platform is secure, accessible, and reliable for businesses that handle sensitive information. "Even with all the advancements in gen-AI, the basis of building an enterprise tool remains the same," says Principal Engineer, Jacques Lemire. "SOC Type 1 and 2 compliance was just the start. We want to make sure game studios can trust their data is always secure and confidential with us."
How Temporal Powers Layer.ai’s Scalability
Layer uses Temporal to streamline their AI-driven workflows for scalable game asset production.
By offloading heavy tasks like asset generation and file processing, Temporal keeps Layer’s servers efficient and with automatic retries, the team can focus on high-impact work. This maintains consistent output at scale, allowing the team to tackle even more complex challenges.
Identifying Technical Challenges
Building a scalable AI-powered tool like Layer came with its own technical challenges. "Generative AI is an inherently complex and ever-evolving space," notes Joona. "There are many permutations of base models, parameters, and training methods. What works today may not work tomorrow, which makes it challenging to keep up with rapid changes."
In their early backend implementation, Layer ran heavy asynchronous tasks like file pre-processing, model inference, and post-processing on the main servers. This caused bottlenecks, slowed user experience, and resulted in an unclear path to handle retries or recover from failures. "Under heavy load, we had frequent task failures, which was a huge issue for maintaining our reliability," Joona adds.
Overcoming Performance Bottlenecks with Temporal
To overcome these challenges, Layer integrated Temporal into their architecture. "By offloading these tasks into Temporal, we were able to decouple them from our main server, preventing overload and improving performance," Joona explains. "Temporal's built-in retries and clear observability gave us better control over task execution, reducing failures and ensuring reliability."
Layer’s AI asset generation pipeline was drastically improved with Temporal’s durable workflows, which guarantee task retries if something goes wrong, ensuring consistency and reliability even under high demand.
Example Architecture
A prime example of how Temporal supports Layer is the ‘FileUploadWorkflow’, which handles the entire file upload process, which a key workflow is Layer’s product, from waiting for uploads to executing post-processing.
This workflow relies on Temporal to ensure that if any step fails (like a timeout during upload), the task is retried automatically, preventing failures from affecting the overall pipeline.
Here’s a snippet of the ‘FileUploadWorkflow’ workflow definition:
@workflow.defn
class FileUploadWorkflow:
TASK_QUEUE = TASK_QUEUE
def __init__(self) -> None:
self.is_file_uploaded = False
@staticmethod
def make_id(id: FileID) -> str:
return f"file-upload-{id}"
@workflow.run
async def run(self, input: FileUploadWorkflowInput) -> FileUploadWorkflowOutput:
self.bucket_path = input.bucket_path
try:
# Wait for the uploaded signal
await workflow.wait_condition(lambda: self.is_file_uploaded, timeout=FILE_PROCESSING_TIMEOUT)
except asyncio.TimeoutError:
# We reached the poll timeout without a signal
file_status = await workflow.execute_activity_method(
ProcessFileActivities.check_file_status,
CheckFileStatusInput(file_id=input.file_id),
start_to_close_timeout=datetime.timedelta(seconds=10),
retry_policy=RetryPolicy(
maximum_attempts=2, initial_interval=datetime.timedelta(seconds=10), backoff_coefficient=1
),
)
if file_status.uploaded:
self.is_file_uploaded = True
if not self.is_file_uploaded:
# Mark file as deleted since it wasn't uploaded within the processing timeout
output = await workflow.execute_activity_method(
ProcessFileActivities.delete_file,
DeleteFileInput(file_id=input.file_id),
start_to_close_timeout=datetime.timedelta(seconds=30),
retry_policy=RetryPolicy(maximum_attempts=5),
)
return FileUploadWorkflowOutput(file_id=output.file_id, success=False)
output = await workflow.execute_activity_method(
ProcessFileActivities.mark_file_uploaded,
MarkFileUploadedInput(file_id=input.file_id),
start_to_close_timeout=datetime.timedelta(seconds=30),
retry_policy=RetryPolicy(maximum_attempts=2),
)
if output.success:
await workflow.execute_child_workflow(
PostProcessFileWorkflow.run,
PostProcessFileWorkflowInput(file_id=input.file_id),
id=PostProcessFileWorkflow.make_id(input.file_id),
task_queue=PostProcessFileWorkflow.TASK_QUEUE,
id_reuse_policy=WorkflowIDReusePolicy.REJECT_DUPLICATE,
)
return FileUploadWorkflowOutput(file_id=output.file_id, success=output.success)
@workflow.signal
async def file_uploaded_signal(self, input: FileUploadedSignal) -> None:
if input.bucket_path == self.bucket_path:
logger.info("Received signal 'file_uploaded'")
self.is_file_uploaded = TrueThis code is a low-level representation of the file upload and processing pipeline. It handles retries, timeouts, and checks to ensure that tasks continue executing, even if certain steps fail.
The diagram below provides a higher-level view of this process, showing how the different components interact within the broader workflow. It visualizes key steps like trigger events, decisions for prompt translation, and the orchestration of tasks like preprocessing, file generation, and post-processing.
As seen in both the code and the diagram, Temporal's retry mechanism and clear visibility ensure that if something goes wrong, the system keeps running smoothly. This decoupling of tasks allows Layer to scale and process complex workflows more reliably.
Redefining the Future of Game Design
Layer.ai is redefining what’s possible in game design, ensuring developers can focus on creating stunning, immersive worlds instead of wrestling with production bottlenecks. It’s the kind of tool that prevents your game from becoming another Rise of the Robots—all hype, no substance—and instead helps you deliver an experience players will love and remember.
Because at the end of the day, cutting-edge tech only matters if it brings your creative vision to life.
To see Layer in action and explore how it can improve your game asset creation process, visit their website and follow them on X and LinkedIn for the latest updates.