re:Invent Guide: For the Games Industry


Serverless Hero

The games industry has matured leaps and bounds in the last decade, and the marketplace has become increasingly crowded. To stand out from the crowd, you need so much more than fancy graphics and smooth gameplay.

Increasingly, top-selling games are starting to personalize the playing experience for individual players in order to reach the widest market. Machine learning (ML) and data analytics are a big part of this process.

In this year’s game tech sessions, we see how the biggest game companies in the world use AWS to help them deliver more fun and excitement for their players.

Recommended Sessions
If you want to see how top-grossing games are using ML to enhance the player experience, check out these sessions.

How Rovio teaches Angry Birds to fly in the cloud using ML ›

Peek behind the curtain, and see how Rovio uses ML to predict and deliver the best levels for Angry Birds players. Its AWS-hosted services process thousands of requests per second and analyze an impressive 4 billion analytics events per day.


How CAPCOM builds fun games fast with containers, data, and ML ›

CAPCOM is one of the most successful video game makers ever, and I have been playing their games for decades, across many generations of hardware. In this session, we get to see the world premiere of a new mobile game they have been developing with the help of AWS services such as Amazon SageMaker, Amazon EKS, and Amazon DynamoDB.


How Call of Duty uses ML to personalize player engagement ›

Activision Blizzard and the Call of Duty franchise has dominated the first-person-shooter (FPS) market for years. Now we get to see how it drives engagement and revenue by personalizing each player’s playing experience with reinforcement learning.

If you want to see what data migration at scale looks like, check out these sessions.

Migrating the live Pokémon database to Aurora PostgreSQL ›

Data migration is the bane of every system migration project. This session is a perfect demonstration of what you should aspire to. Check out how Pokémon Go migrated 300 million players’ data to Amazon PostgreSQL and Amazon DynamoDB with zero downtime.

If you want to dive deep into specific technologies and have a live debate, check out this chalk talk.

Game-changing analytics: Know your players, grow your game ›

I always like chalk talks; they’re less rigid and much more interactive. You get to dive deep into a particular solution and understand why certain technology choices are made. And best of all, you can bring your own challenges to the table and get feedback from the experts on how you might be able to apply these technologies. In this session, we focus on analytics pipelines with Amazon S3, Amazon Kinesis, and Amazon Athena.

If you want to get your hands dirty and learn how to build a lifelike AI for your game, check out this session.

Land a rocket (and a good game) with reinforcement learning ›

This is a two-hour workshop that gives you hands-on experience with reinforcement learning using Amazon SageMaker RL. You learn how to train a lunar lander vehicle in a Box2D simulation to land itself on the surface of the moon. And you get to compete against other teams to see who can create the best model!

Here are a few more sessions that are not in the game tech session collection, but are still relevant.

Build cost-effective web & containerized apps on EC2 Arm instances ›

This is another chalk talk, which means diving deep into a real-world application. In this case, NetEase Games (a well-known game developer in China) shares its experience with the Amazon EC2 A1 instance type. They talk about how they use these Arm-based instances to run their large-scale online gaming platform and achieve significant cost savings.


Extend analytics beyond your data warehouse ›

Learn how to extend SQL queries across data in Amazon Redshift and Amazon S3 without wasting time with complex ETL processes to load your data. This solves the common problem of dark data—data that is often not analyzed due to the complexity and cost of loading it into traditional data warehouse solutions.


AWS analytics enables fraud prevention for Sony’s PlayStation ›

See how Sony Interactive Entertainment built its fraud prevention system with Amazon Kinesis Data Analytics, Amazon Kinesis Data Firehose, AWS Lambda, AWS Glue, Amazon S3, and Amazon DynamoDB. The system substantially increased the approval rate for user transactions without the usual drawbacks such as increased declines and chargebacks from banks. Talk about having your cake and eating it too!


Delight your customers with ML-based personalized recommendations ›

In this session, you see how Voodoo Games (creator of hits such as Helix Jump and Fire Balls 3D) uses Amazon Personalize to create personalized recommendations and increase engagement and revenue.

This year’s re:Invent has so much to offer. If the sessions I have recommended here piqued your interest in AI and ML, then you should check out other great sessions in AI and ML. Look for sessions with “AIM” in the ID in the session catalog.

Also, don’t forget to visit the AWS Game Tech Game Zone in the Venetian, Level 4, East Alcove. You can play games there and learn about the AWS services that power them. It’s a great way to network with industry peers and meet the AWS Game Tech team. There is also a special game tech networking event on Tuesday, Dec. 3. You can ask your AWS account manager to get an invite.

But most of all, enjoy re:Invent! I hope you have a wonderful time there.

Explore other AWS Hero Guides ›

Registration now open

Register now