Benefits
Games Created
Total Game Plays
Overview
SEELE combines a state-of-the-art (SOTA) multimodal Large Language Model (LLM) with a gamified Integrated Development Environment (IDE). Creators build playable 3D worlds with conversations and simple prompts, without writing code. SEELE generates code, models, textures, and audio instantly to deliver fully playable games. Currently, the platform hosts over 366,100 active creators who have built more than 200,000 games. SEELE uses core Amazon Web Services (AWS), including Amazon Bedrock（for overseas business only） and Amazon EKS .
Opportunity
Reinventing Traditional Game Development with Agentic AI
As AI technology reshapes industrial value chains, the 3D content generation field undergoes an efficiency revolution. SEELE, as an AI-native company, uses AI to transform how 3D games are created, distributed, and played. SEELE integrates Agentic AI across product research and development (R&D), game generation, and asset production. As the business grows rapidly, SEELE requires a more stable service foundation to support global expansion in the future.
After evaluating various cloud providers, SEELE chose AWS for its core advantages, optimizable billing, global infrastructure, and robust security and compliance. These features preserve SEELE's technical neutrality and architectural freedom while reducing long-term risks and supporting sustainable growth at scale.
Solution
Driving a "Double Reduction" in Costs and Operational Workload with AWS
During game generation, SEELE's core requirements for Agentic AI are autonomous task decomposition and external tool invocation. For overseas business scenarios, AWS uses core services including Amazon Bedrock and Amazon EKS to optimize both cost and workflow for SEELE:
- Build Automated Solutions: Using the Claude model in Amazon Bedrock, SEELE builds end-to-end automation. With guidance from AWS architects, SEELE optimizes cache point settings, reducing token consumption and text-processing costs.
- Optimize Deployment and Operations: SEELE uses Amazon EKS clusters to effectively resolve core pain points in large-scale deployment and operations. Amazon EKS Auto Mode handles traffic spikes, reduces operational workload, and improves operational efficiency.
By building our Agentic AI architecture on AWS, we implemented a deep transformation from Infrastructure Migration to Intelligence-Driven Operations. This significantly reduces costs while multiplying the efficiency of core decision-making.
Wang Shimu
Founder and CEO, SEELEOutcomes
Accelerating Global Expansion of Agentic AI Game Creation Tools on a Robust Cloud Foundation
SEELE now hosts all existing services on AWS. Compared to the previous deployment model, migrating to Amazon EKS fundamentally transforms the research and development model of SEELE. The team adopts DevOps and GitOps to improve development efficiency and service stability. The Amazon EKS-based deployment strategy also reduces operational costs significantly, accelerating the global expansion process.
Looking ahead, SEELE plans to deepen its collaboration with AWS in global network acceleration, automatic database cluster synchronization, and multi-region active-active deployment. These initiatives will support highly available, low-latency, and responsive services.
About SEELE
SEELE is the only platform that combines a SOTA multimodal LLM with a gamified IDE. On SEELE, creators generate code, models, textures, and audio without starting from a blank canvas or having professional programming or modeling skills. From pixel platformers to physics puzzles, users select a proven game archetype, and AI builds the core gameplay loop in seconds, delivering a playable game.
Disclaimer: Amazon Web Services currently deploys the aforementioned certain generative AI-related services in Global regions. Amazon Web Services China region services are operated by NWCD and Sinnet, with more details at the official Amazon Web Services China region website.
