Hello everyone. Welcome to Con 42, rust Plan 2025. My name is Vin Kana and I'm thrilled to be here to talk about something I have been personally passionate about combining the power of Rust with AWS Cloud services. Over the next several minutes, we will explore how rust performance, safety, efficiency, make it a great choice for building cloud native applications that scaled reliably instead of just theory, we'll walk through our flowing. Example system step by step. So you live with a clear mental model of how this can apply to a real world. Workloads when you build on AWS scale is always central applications may need to handle millions of events, thousands of concurrent requests and unpredictable workloads at that level, every milliseconds of latency and every MB of memory counts. Rust combines predictable performance similar to c RC plus with compiled and guarantees that eliminates common runtime issues for Lambda functions or ECS tasks. Small binaries reduce cold start star types and lean memory uses cut cuts costs. This blend of performance and safety makes rust an excellent choice when you care about efficiency and reliability. Let's compare rust agonist to other popular languages. In AWS Go is a strong for Lambda and containers, thanks to go routines, but it, its runtime is bigger than rust. Java has massive ecosystem, but coal starts and memory overheads are costly. No JS offers fast development, but struggles with CPU bound workloads. Python is unbeatable for prototyping and ML libraries. Though it slows down at scale, c plus delivers raw power, but manual memory makes it error prone rust balances performance near c plus with safety and compiled and guarantees that combine that combination makes rust stand out for cloud workloads. Where does rust shame for Lambda? Small binaries equal for or equal faster cold starts. The borrow checker enforces memory safety, preventing bugs before develop deployment. With Tokyo, you can run thousands of async calls to AWS services without overwhelming threats. That reduces CPU and memory lowering costs In loan Lambda, ECS and EKS, strong typing issues, requests are valid at compiled, preventing misconfigurations from ever reaching production. But rust isn't perfect. Compiled times can be long. In large projects, ownership and lifetimes have a steep learning curve. The ecosystem is maturing, but Java and Python have more polished libraries. Tooling for Lambda is improving, but Python and Node still offer more templates and rust favors explicitness over dynamic meta programming. Which is safer but less flexible for quick experiments. Being aware of these trade offs makes adoption decision realistic. The A-W-S-S-D-K for Rust provides strong support. Each service is its own create, so you input only what you need. Shared configuration handles, credentials and regions securely built on Tokyo. It's first making it. Efficient for input output. With retries, ation and streaming built in, developers can focus on value rather than boiler plate. To make this concrete, let's build a file ingestion pipeline. A user uploads to S3 and even flows through event bridge into SQS. Rust worker workers process, it DynamoDB storage results and observability keeps watch. This example matters. Real world patterns and highlights. Russ value at each step. Step one, upload files. With Russ STK, you gain compiled time. Assurance that requests are valid, I think lets you upload many files concurrently without blocking. This builds confidence that what you send to AWS S3 is correct and efficient. Next triggering process. C, even Bridge sends S3 events into SQS. Trust workers consume messages safely DC realizing them with third, A sync workers scale out efficiently, letting you process hundreds of events in parallel without adding threats. This is where Rust, A sync runtime shines. For compute, we'd apply to Lambda Rust. Binaries are small, giving faster cold stocks. Execution is predictable and efficient. With As think a single function can manage multiple tasks compared to Java or T net Runtimes, rust Lambda are lighter, cheaper, and faster After processing, we persist results in Dynamo DB Rust. Builders enforce correctness at compiler as think queries let you scale to thousands of operations. The system stays responsive. Even under heavy load observability is essential. With the tracing crate, you add structured spans into a sync workflows pair with open telemetry and metrics stream into CloudWatch for errors. Anyhow, are the errors capture detailed context together, this makes debugging faster and operations. No, AWS system is complete Without security and deployment use IAM with the least privileged role. And KMS to encrypt sensitive data for Lambda Cross compiled to muzzle for small static binaries in containers. Multi-stage builds, strip unnecessary dependencies, shrinking image size, and improving security. To recap, rust compile time. Safety prevents runtime bug, I think first makes it perfect for input output workload. The modular SDK is easy to adapt and end-to-end patterns map naturally compared to other languages. Rust blend of performance and safety is unique. In closing, rust is ready for cloud. It combines system level performance with strong safety guarantees, makes making it ideal for AWS workloads at scale. The SDK provides the tools and our example showed rust across AWS landscape. If you're building next generation cloud native apps, rust is not just experimental, it's production ready. Thank you and I hope this inspires you to try Rust in your AWS projects. I.