Constructing clever brokers to deal with complicated real-world duties will be difficult. Moreover, organizations typically must fine-tune and customise smaller, specialised fashions to carry out higher in particular use circumstances, reasonably than relying solely on giant, pre-trained fundamental fashions. The AWS AI League presents progressive applications to assist corporations overcome the challenges of constructing superior AI capabilities by thrilling competitions that drive innovation in agent AI and mannequin customization.
In 2025, the primary AWS AI League competitors attracted the eye of builders, information scientists, and enterprise leaders world wide. They got here collectively to resolve urgent issues utilizing the most recent AI instruments and methods. The grand finale of AWS re:Invent 2025 was an thrilling showcase of their ingenuity and abilities. Cross-functional groups from main organizations competed head-to-head to display their capacity to create efficient prompts, fine-tune fashions, and construct highly effective AI brokers.
Congratulations to our 2025 AWS AI League Champions! After a fierce competitors between these three excellent builders, we received and shared a $25,000 prize pool.
- 1st Place: Hemanth Vediyera from Cisco
- 2nd place: Ross Williams (Akufar)
- third place: Deepesh Khanna of Capital One
Determine 1: From left to proper: Ross, Hemanth, Deepesh
This submit describes learn how to use the AWS AI League program to host an AI competitors. This permits members to expertise the ideas of mannequin customization and agent constructing, apply these to sort out real-world enterprise challenges, and showcase progressive options in an enticing game-style format. Give attention to new agent AI and mannequin customization challenges. Companies can apply to host inner tournaments utilizing AWS credit, and builders can compete at AWS occasions.
To get began, go to the AWS AI League product web page.
What’s the AWS AI League Championship?
Your AWS AI League expertise begins with a 2-hour hands-on workshop with an AWS skilled, adopted by experiments at your personal tempo. The journey culminates in a fascinating sport show-style grand finale. There, we showcase AI creations and options to deal with urgent enterprise challenges. The next diagram exhibits these three steps.
Determine 2: AWS AI League Championship Steps
Constructing on the success of our 2025 program, we’re excited to announce the launch of the AWS AI League 2026 Championship. This yr’s competitors options two new challenges that permit members to place their AI abilities to the check.
- Agentic AI Problem lets you construct clever brokers utilizing Amazon Bedrock AgentCore. Opponents are creating personalized agent architectures to sort out real-world enterprise issues.
- The Mannequin Customization Problem, which enhances the Agent AI Problem, now makes use of SageMaker Studio’s newest tweak recipes. Now you’ll customise the mannequin on your particular use case.
The 2026 AI League Championship will double the prize pool to $50,000 and have tracks for builders of various ability ranges, from newbie to superior.
Construct clever brokers with the Agent AI Problem
AWS AI League now consists of thrilling Agent AI Challenges. On this problem, you employ Amazon Bedrock AgentCore to construct clever brokers to resolve complicated issues in dynamic game-style competitors. On this problem, brokers navigate a maze-like grid atmosphere and encounter varied challenges whereas looking for treasure chests. These challenges correspond to real-world use circumstances and check the agent’s capacity to deal with inappropriate content material, execute code, use a browser, and extra.
Brokers have a time restrict to traverse the map, acquire factors, and overcome obstacles earlier than reaching the treasure chest. The extra factors you earn, the upper you rank on the leaderboard. Amazon Bedrock AgentCore primitives can help you absolutely customise your brokers, permitting you to extra securely scale and handle production-grade brokers. It’s also possible to select particular fashions for supervisors and subagents, or create customized instruments similar to Bedrock Guardrails, AgentCore Reminiscence, or AWS Lambda features to assist your brokers deal with your challenges. The next diagram exhibits the obstacles that the agent should overcome whereas touring to achieve the treasure chest.
Determine 3: AWS AI League Agent Problem
AWS AI League offers an entire consumer interface (UI) for customers to construct clever agent options. You should use this no-code UI to construct multi-agent architectures and instruments, integrating varied elements similar to Amazon SageMaker Studio CodeEditor for interactive coding of customized Lambda features and instruments. This lets you develop and customise agent-based options fully throughout the AWS AI League web site with out leaving your atmosphere.
The next screenshot exhibits the agent constructing expertise throughout the AWS AI League web site.
Determine 4: AWS AI League agent software
Determine 5: AWS AI League multi-agent structure
Via contests, customers obtain real-time efficiency suggestions on their brokers, and large-scale language mannequin (LLM) evaluators present scores to assist iterate. The next picture exhibits how brokers are evaluated throughout a problem.
Determine 6: AWS AI League Agent Problem Analysis
On the grand finale, the highest finalists will take to the stage to showcase their agent skills in a stay sport present format, demonstrating the ability and flexibility of agent AI in fixing complicated multi-step issues. Analysis standards embrace time effectivity, drawback fixing accuracy, agent planning, and token consumption effectivity. The next snapshot exhibits the ultimate spherical of the grand finale of re:Invent 2025.
Determine 7: AWS AI League re:Invent 2025 Grand Finale
Customise your mannequin to carry out higher than bigger fashions
The AWS AI League is increasing the scope of mannequin customization challenges and enabling you to make the most of the most recent advances in fine-tuning methods.
You’ll be able to entry a brand new mannequin customization expertise and use highly effective new coaching recipes inside Amazon SageMaker Studio. The aim is to develop extremely efficient domain-specific fashions that outperform large-scale reference fashions.
This problem begins by honing your mannequin customization abilities. Utilizing the instruments and methods you have discovered, apply superior fine-tuning strategies to enhance mannequin efficiency. After customizing the mannequin, the precise testing begins. The mannequin is submitted to a leaderboard for efficiency analysis in comparison with the reference mannequin. The mannequin earns factors every time automated selections decide that the personalized mannequin’s response is extra correct and complete than the reference mannequin’s output. Showcase your superior abilities and rise to the highest of the leaderboard, doubtlessly opening up new alternatives on your group.
Throughout a problem, you undergo a leaderboard and obtain real-time suggestions in your mannequin’s efficiency from an automatic evaluator. Leaderboards consider submissions towards reference datasets all through the competition, offering instantaneous suggestions on accuracy to assist iterate and enhance your answer. The next picture exhibits how AI critique is used to guage personalized fashions.
Determine 8: Custom-made analysis of the AWS AI League mannequin
On the grand finale, the highest finalists will display their mannequin’s capabilities in a stay sport present format and exhibit their fast engineering skills. Scoring in the course of the sport present consists of skilled scores through which area consultants and stay viewers take part in real-time voting to find out which AI options finest clear up real-world enterprise challenges. The next picture exhibits the participant’s immediate engineering view in the course of the grand finale.
Determine 9: AWS AI League Mannequin Customization Grand Finale Participant View
conclusion
On this submit, we explored the AWS AI League’s new challenges and the way they’re remodeling the best way organizations method AI improvement. At AWS, we have discovered that the quickest technique to foster innovation is thru competitors. AWS AI League permits builders to showcase their AI abilities, compete, and innovate.
To be taught extra about internet hosting an AWS AI League in your group, go to AWS AI League and be taught extra about constructing clever brokers and customizing AI fashions. AWS AI Training Catalog above AWS Skill Builder.
Concerning the creator
mark karp I am an ML Architect on the Amazon SageMaker Service workforce. He focuses on serving to clients design, deploy, and handle ML workloads at scale. In my free time, I take pleasure in touring and exploring new locations.
Natasha Ok. Idris I’m a Product Advertising and marketing Supervisor for the AWS AI/ML Gamified Studying Program. She is obsessed with democratizing AI/ML abilities by partaking, hands-on instructional initiatives that bridge the hole between superior know-how and real-world enterprise implementation. Her experience in constructing studying communities and driving digital innovation continues to form our method to creating impactful AI education schemes. Exterior of labor, Natasha enjoys touring, cooking Southeast Asian delicacies, and exploring nature trails.
