AWS re:Invent 2025 was one other unimaginable occasion! Tens of 1000’s of builders got here collectively to study concerning the newest AWS improvements. For the ElastiCache crew, it was a priceless alternative to fulfill a number of clients and listen to straight from them about what’s working effectively and what challenges they’re going through – insights that straight form our roadmap to raised serve them. On the identical time, we acquired to showcase new ElastiCache capabilities that assist clients unlock even better efficiency and worth.
The important thing themes this yr have been upgrading to Valkey from Redis, caching methods for generative AI and agentic AI workloads, and constructing cost-efficient, high-performance purposes with ElastiCache Serverless. In case you missed a few of these classes, otherwise you wished to get caught up on why clients like Expedia, Scopely, Adobe, and Amazon.com are constructing on ElastiCache, you’ll be able to learn this beneficial abstract of a number of the ElastiCache highlights from re:Invent 2025.
Management Session: Databases made easy so brokers and builders can change the world
From startups to international enterprises, AWS Databases has been the trusted basis powering revolutionary purposes for practically twenty years. Be a part of G2 Krishnamoorthy, Vice President of AWS Database Providers, as he presents our mission to remodel the way you handle your knowledge. Uncover our imaginative and prescient the place database administration turns into easy, delivering sturdy efficiency, seamless scale, and complete safety. Learn the way databases combine with AI brokers and instruments to cut back operational complexity whereas supporting a brand new era of AI-native purposes to advance what you are promoting capabilities. Hear from Vercel and Robinhood as they construct new alternatives with AWS.
For ElastiCache particularly, this session covers semantic caching with ElastiCache for Valkey, which achieves the bottom latency vector search with highest throughput and greatest price-performance at 95% recall charge amongst fashionable vector databases on AWS for each agentic AI (equivalent to conversational reminiscence) and generative AI (equivalent to Retrieval Augmented Era) use instances. The session additionally covers how frameworks like Mem0, LMCache, and LangGraph use ElastiCache for Valkey to attain millisecond retrieval efficiency for short-term agentic reminiscence.
To study extra about semantic caching with ElastiCache, see Decrease value and latency for AI utilizing Amazon ElastiCache as a semantic cache with Amazon Bedrock.
Scaling MONOPOLY GO! to thousands and thousands with DynamoDB & ElastiCache
Scopely shared their unimaginable journey of scaling MONOPOLY GO! to serve thousands and thousands of concurrent customers with minimal assets. Learn the way they used Amazon DynamoDB and ElastiCache Serverless to energy MONOPOLY GO!’s speedy progress, dealing with an unimaginable 2.1 million writes per second at peak visitors. Uncover the architectural patterns and optimization methods that enabled seamless matchmaking, good friend discovery, and real-time gameplay for thousands and thousands of concurrent customers. Discover sensible insights on managing unpredictable visitors spikes, decreasing operational complexity, and reaching value financial savings by means of serverless applied sciences.
Higher, sooner, cheaper: How Valkey is revolutionizing caching
Uncover how the highly effective mixture of open supply Valkey and Amazon ElastiCache is reworking the panorama of distributed caching. Discover latest improvements equivalent to probabilistic knowledge buildings like bloom filters, a reminiscence environment friendly dictionary that reduces reminiscence utilization by as much as 40%, and a multi-threaded structure that helps instantaneous scaling by as much as 230%. Study concerning the infrastructure that permits ElastiCache to seamlessly patch a whole lot of 1000’s of consumers whereas sustaining excessive availability, even for excessive workloads like Prime Day’s trillion requests per minute.
Disagree in Commits: The Efficiency Enhancements That Reduce Prices by a Third
When paths diverge in open supply, innovation typically follows. Be a part of Corey Quinn and Madelyn Olson as they dive into how Amazon ElastiCache for Valkey reduce costs by as much as 33% whereas concurrently bettering each efficiency and reliability by means of community-driven improvement. You’ll uncover how open governance allows optimization selections that profit all customers, why efficiency enhancements and stability aren’t mutually unique, and what it actually takes to create a seamless migration path at scale. Whether or not you’re presently operating Redis, evaluating options, or just eager about how open-source economics can profit cloud clients, you’ll go away with sensible insights on decreasing infrastructure prices with out compromising on reliability.
Optimize agentic AI apps with semantic caching in Amazon ElastiCache
Multi-agent AI techniques now orchestrate advanced workflows requiring frequent basis mannequin calls. Learn the way semantic caching with Amazon ElastiCache for Valkey reduces latencies from single-digit seconds to single-digit milliseconds whereas slicing basis mannequin prices for manufacturing workloads. By implementing semantic caching in agentic architectures like RAG-powered assistants and autonomous brokers, you’ll be able to create performant and cost-effective production-scale agentic AI techniques.
Construct serverless chatbots utilizing Amazon ElastiCache & Aurora PostgreSQL
Learn to construct a serverless-first chatbot utility by leveraging Amazon Aurora with pgvector as a vector retailer and Amazon ElastiCache Serverless for Valkey to handle chat historical past, semantic caching, and agent reminiscence. Uncover how one can use these providers to create contextually conscious, scalable chatbots with agentic capabilities that retain conversational reminiscence throughout interactions. Discover architectural patterns, implementation methods, and real-world use instances. Achieve insights into optimizing question latency, efficiency, and cost-effectiveness. This session equips you with sensible data to construct fashionable, serverless-powered chatbot purposes utilizing generative AI.
Deep dive into AWS Database Financial savings Plans
Database Financial savings Plans are a versatile pricing mannequin that reduces your database prices by as much as 35% once you decide to a constant quantity of utilization over a 1-year time period. This financial savings plan robotically applies to eligible serverless and provisioned occasion utilization no matter engine, occasion household, measurement, deployment possibility, or AWS Area with no upfront cost. For instance, with Database Financial savings Plans, you’ll be able to change between occasion sorts equivalent to Aurora db.r7g and db.r8g situations, shift a workload from EU (Eire) to US (Ohio), or migrate from ElastiCache for Valkey situations to ElastiCache Serverless for Valkey and nonetheless profit from the discounted charges. The session covers utilization sample evaluation, value optimization methods, and how one can maximize financial savings whereas sustaining workload flexibility.
For ElastiCache customers, Database Financial savings Plans apply to each ElastiCache Serverless for Valkey (30% low cost) and ElastiCache for Valkey situations (20% low cost).
Optimizing for prime efficiency with Amazon ElastiCache Serverless
Learn to construct extremely scalable, high-performance purposes with Amazon ElastiCache Serverless. This session reveals how ElastiCache Serverless works behind the scenes and offers metrics that can assist you determine and debug utility efficiency points. Uncover how one can construct efficient purposes with ElastiCache Serverless and use the design of the service to your benefit.
Superior knowledge modeling for Amazon ElastiCache
This session delves into the intricacies of Amazon ElastiCache knowledge modeling utilizing purpose-built Valkey knowledge sorts to optimize utility efficiency and scalability. Discover the usage of strings, units, sorted units, hashes, bitmaps, and geospatial indexes to mannequin advanced relationships and remedy use instances equivalent to caching, session retailer, function retailer, real-time analytics, geospatial purposes, and charge limiters.
Video sampling & search utilizing ElastiCache & multimodal embeddings
You need to use multimodal embeddings for a variety of video evaluation capabilities, from content material understanding and summarization to classification and suggestions. Learn to use good sampling and video search, producing multimodal embeddings in Amazon Bedrock and utilizing vector search powered by Amazon ElastiCache. See stay coding examples to find how one can construct pleasant video experiences on your customers.
Key highlights
Three key highlights from this yr’s ElastiCache classes:
- Serverless at scale: Prospects are more and more upgrading their workloads to ElastiCache Serverless. MONOPOLY GO! achieved 2.1M writes/second with ElastiCache Serverless, demonstrating how serverless can deal with workloads of nearly any scale.
- Valkey delivers: By upgrading to ElastiCache for Valkey, clients can obtain as much as 40% reminiscence discount, 230% scaling enhancements, and 33% lowered pricing. Prospects can improve from ElastiCache for Redis to ElastiCache for Valkey with nearly no downtime.
- Function constructed semantic caching: ElastiCache delivers lowest latency vector search with highest throughput and greatest price-performance at 95% recall charge amongst fashionable vector databases on AWS. With vector search, clients can cut back latencies from single-digit seconds to single-digit milliseconds whereas slicing basis mannequin prices for manufacturing workloads.
- Valkey as a reminiscence layer for brokers: Newly launched connectors enable clients to make use of ElastiCache for Valkey for short-term agentic reminiscence by means of frameworks equivalent to Mem0, LMCache, and LangGraph.
- Database Financial savings Plans: Prospects can use Database Financial savings Plans to cut back database prices by as much as 35% with a 1-year dedication and no upfront cost required. Database Financial savings Plans can be found for ElastiCache for Valkey.
Getting began
To study extra about Amazon ElastiCache and get began, go to the ElastiCache product web page. For Valkey particular steering, take a look at What’s Valkey?
We’re excited concerning the improvements showcased at re:Invent 2025 and look ahead to seeing what you construct with ElastiCache. Thanks for becoming a member of us at re:Invent 2025, and we will’t wait to see you at re:Invent 2026!
In regards to the authors
