Building Cloud Messaging Architecture With AWS

Your AI chatbot forgets everything when a user switches between messaging platforms. Here's how to fix that. Most chatbots treat each channel as a separate world. A user shares a photo on WhatsApp, then asks about it on Instagram. The agent has no idea what they're talking about. I built a multichannel AI agent that maintains persistent memory across messaging platforms using Amazon Bedrock AgentCore. One deployment, shared identity, full context. The demo uses WhatsApp and Instagram, but the architecture extends to any messaging channel: Slack, Telegram, Discord, SMS. How it works: → Unified identity: deterministic user IDs per channel (wa-user-{phone}, ig-user-{sender_id}) mapped to a single actor in AgentCore Memory. Adding a new channel means adding one more ID pattern. → Two memory layers: short-term (conversation turns with TTL) and long-term (extracted facts, preferences, summaries that persist indefinitely) → Multimodal processing: text, images (Claude vision), voice (Amazon Transcribe), video (TwelveLabs), and documents → Smart buffering: DynamoDB Streams with 10-second tumbling windows batch rapid messages before invoking the agent The architecture uses three AWS CDK stacks: Stack 00 → AgentCore Runtime + memory layer Stack 01 → WhatsApp (AWS End User Messaging) or Stack 02 → Multi-channel API Gateway (WhatsApp + Instagram + any new channel) Users can even link their accounts across platforms through conversation. The agent merges identities in a unified DynamoDB table. The core memory and identity layers are channel-agnostic. WhatsApp and Instagram are the first two integrations, but the pattern is designed to grow. Full code and deployment guide are open source. Each stack deploys in about 15 minutes #AI #Chatbot #Agents #AWS #LLM

Modern financial operations demand the ability to process millions of invoices daily, with low latency, high availability, and real-time business visibility. Traditional monolithic systems struggle to keep up with the surges and complexity of global invoice processing. By adopting an event-driven approach, organizations can decouple their processing logic, enabling independent scaling, real-time monitoring, and resilient error handling. Amazon Simple Queue Service (#SQS) and Amazon Simple Notification Service (#SNS) enable resilience and scale in this architecture. SNS acts as the event router and broadcaster in this architecture. After events are ingested (via API Gateway and routed through EventBridge), SNS topics are used to fan out invoice events to multiple downstream consumers. Each invoice status—such as ingestion, reconciliation, authorization, and posting—gets its own SNS topic, enabling fine-grained control and filtering at the subscription level. This ensures that only relevant consumers receive specific event types, and the system can easily scale to accommodate new consumers or processing requirements without disrupting existing flows. Each SNS topic fans out messages to one or more SQS queues. SQS provides the critical function of decoupling the event delivery from processing. This means that even if downstream consumers (like AWS Lambda functions or Fargate tasks) are temporarily overwhelmed or offline, no events are lost—SQS queues persist them until they can be processed. Additionally, SQS supports dead-letter queues (DLQs) for handling failed or unprocessable messages, enabling robust error handling and alerting for operational teams. Specific to resilience and scale, look at these numbers.... • Massive Throughput: SNS can publish up to 30,000 messages per second, and SQS queues can handle 120,000 in-flight messages by default (with quotas that can be raised). This supports surges of up to 86 million daily invoice events. • Cellular Architecture: By partitioning the system into independent regional “cells,” each with its own set of SNS topics and SQS queues, organizations can scale horizontally, isolate failures, and ensure high availability. • Real-Time Monitoring: The decoupled, event-driven flow—powered by SNS and SQS—enables near real-time dashboards and alerting, so finance executives and auditors always have up-to-date visibility into invoice processing status. #financialsystems #cloud #data #aws https://lnkd.in/gNnYpeu7

𝗨𝗻𝗹𝗼𝗰𝗸𝗶𝗻𝗴 𝗦𝗰𝗮𝗹𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝘄𝗶𝘁𝗵 𝗘𝘃𝗲𝗻𝘁-𝗗𝗿𝗶𝘃𝗲𝗻 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 𝗼𝗻 𝗔𝗪𝗦 Tired of tightly coupled systems that crumble under scale?  𝗘𝘃𝗲𝗻𝘁-𝗗𝗿𝗶𝘃𝗲𝗻 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 (𝗘𝗗𝗔) might be your savior. By leveraging AWS services like 𝗦𝗡𝗦, 𝗦𝗤𝗦, 𝗘𝘃𝗲𝗻𝘁𝗕𝗿𝗶𝗱𝗴𝗲, and 𝗟𝗮𝗺𝗯𝗱𝗮, you can build systems that scale seamlessly, react in real-time, and stay resilient.  Let’s dive in! 𝗕𝗲𝘀𝘁 𝗣𝗿𝗮𝗰𝘁𝗶𝗰𝗲𝘀 𝗬𝗼𝘂 𝗖𝗮𝗻’𝘁 𝗜𝗴𝗻𝗼𝗿𝗲: 1️⃣ 𝗗𝗲𝗰𝗼𝘂𝗽𝗹𝗲 𝗖𝗼𝗺𝗽𝗼𝗻𝗲𝗻𝘁𝘀: Use SNS (pub/sub) and SQS (message queues) to separate producers and consumers. No more bottlenecks! 2️⃣ 𝗘𝘃𝗲𝗻𝘁𝗕𝗿𝗶𝗱𝗴𝗲 𝗳𝗼𝗿 𝗘𝘃𝗲𝗻𝘁 𝗕𝘂𝘀𝗲𝘀: Centralize events across services (SaaS, AWS, custom apps) for cleaner integration. 3️⃣ 𝗟𝗮𝗺𝗯𝗱𝗮 𝗳𝗼𝗿 𝗦𝗲𝗿𝘃𝗲𝗿𝗹𝗲𝘀𝘀 𝗖𝗼𝗺𝗽𝘂𝘁𝗲: Trigger functions in response to events without managing servers. Think: Cost-efficient scaling. 4️⃣ 𝗣𝗿𝗶𝗼𝗿𝗶𝘁𝗶𝘇𝗲 𝗘𝗿𝗿𝗼𝗿 𝗛𝗮𝗻𝗱𝗹𝗶𝗻𝗴: Use SQS dead-letter queues (DLQs) to capture failed events and retry logic. 5️⃣ 𝗠𝗼𝗻𝗶𝘁𝗼𝗿 𝗘𝘃𝗲𝗿𝘆𝘁𝗵𝗶𝗻𝗴: Pair CloudWatch with X-Ray to trace event flows and debug faster. 𝗥𝗲𝗮𝗹-𝗪𝗼𝗿𝗹𝗱 𝗨𝘀𝗲 𝗖𝗮𝘀𝗲𝘀: ✅ 𝗘-𝗖𝗼𝗺𝗺𝗲𝗿𝗰𝗲 𝗢𝗿𝗱𝗲𝗿 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴: • Order service → SNS → Lambda (inventory check) + SQS (payment processing). • Scalable even on Black Friday! ✅ 𝗜𝗼𝗧 𝗗𝗮𝘁𝗮 𝗜𝗻𝗴𝗲𝘀𝘁𝗶𝗼𝗻: • Thousands of devices → EventBridge → Lambda (transform data) → S3/DynamoDB. • Handle spikes without breaking a sweat. ✅ 𝗥𝗲𝗮𝗹-𝗧𝗶𝗺𝗲 𝗡𝗼𝘁𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀: • User action → SNS → Lambda (send email/SMS) + SQS (audit logs). • Async, fault-tolerant, and fast. 💡 𝗪𝗵𝘆 𝗜𝘁 𝗠𝗮𝘁𝘁𝗲𝗿𝘀: EDA isn’t just a buzzword; it’s a blueprint for systems that adapt to demand, reduce dependencies, and cut costs. Whether modernizing legacy apps or building cloud-native solutions, AWS’s event-driven toolkit has you covered. 𝗬𝗼𝘂𝗿 𝗧𝘂𝗿𝗻: Have you implemented EDA on AWS? #AWS #awscommunity #EventDrivenArchitecture

Should we use SNS, or just use Postgres as our messaging system? Save this for later, when its time to make the choice. The answer is nuanced. Both can be appropriate and can become operational nightmares when used for the wrong workload. Here’s the tradeoff matrix we’ve learned building low-latency durable notifications on Postgres: SNS strengths: - Massive fanout scalability - Managed infrastructure - Easy broadcast to many consumers - Integrates naturally with AWS ecosystems SNS is excellent when events are independent, asynchronous and eventually consistent. But SQS pushes complexity upward: - delivery guarantees - ordering - retries - deduplication - consumer coordination - replay semantics - transactional consistency with your DB The business state and the messaging state live in separate worlds. You often end up rebuilding workflow semantics in application code. Postgres strengths - State + messaging in one transaction - Strong consistency guarantees - Atomic writes and notifications - Simpler observability/debugging - Natural fit for workflow orchestration For many production systems, correctness and coordination matter more than infinite fanout. But Postgres messaging also has its limits: for example, the DB resources are shared and it requires careful indexing. My rule of thumb: if I need raw speed and have 0 consistency requirements, I'll just use SNS. Otherwise, Postgres gives me a much stronger architecture. #postgres

300+ million daily active users Sharing 5+ billion snaps a day This is how Snapchat manages processes in the cloud 👇 Snapchat largely depends on AWS to support its platform on cloud infrastructure, enabling operational efficiency and improving scalability from time to time. Sending a Snap ➡️ - Snap Transmission: After sending a snap from Friend 1 to Friend 2 (on iOS or Android devices), the Snapchat Gateway taps into the Elastic Kubernetes Service (EKS), creating a virtual environment for the swift management of containerized services. - After entering the Gateway, the snap first talks to the Media delivery service and sends the snap to the CloudFront. CloudFront speeds up the distribution of image files. - From CloudFront, the snap persists in the S3 bucket, so it's closer to the recipient when needed. Orchestration 🤖 - Within the Snapchat infrastructure, the MCS is the core orchestration service. After the media is stored in S3, a request is made to MCS. - Given Snapchat's emphasis on securely connecting with 'close friends,' the Friend Graph verifies that the sender has permission to send the message to the recipient. - If the Friend Graph gives a green signal to the snap, all the conversation metadata is persisted in Snapchat’s own database, called SnapDB. The team uses SnapDB as the frontend with DynamoDB as the backend. Why DyanmoDB? - DynamoDB enables massive scalability. Snapchat lets them store 400TB of data. - Additionally, the engineering team developed advanced features such as transactions, TTL, and efficient handling of ephemeral data in their custom layer. - The team manages incremental state synchronization not to put a lot of burden on DynamoDB. This is also how they keep their database costs under control. - Snapchat also does nightly runs on DynamoDB. It runs 2 billion rows per minute to do various things like looking for friend suggestions or deleting ephemeral data. The Receiving End 📲 - Receiving the snaps is latency-sensitive because the friend needs to receive the snap ASAP. To push the message, the messaging service looks up a connection ID/Server ID from Elasticache to get access to a persistent connection that the server has to the client. - Snapchat looks up the metadata from Elasticache. It finds the server the client is already on, and through the Gateway, it sends the snap to the recipient user. To give a bigger picture, there are 900+ such EKS clusters running, and each cluster has more than 1000 instances. In managing 900+ EKS clusters, Snapchat achieves: ➔ Scalability: Geared for 300+ million daily users. ➔ Enhanced User Experience: Reducing snap latency by 24%. ➔ Cost Efficiency: Leveraging auto-scaling and Graviton instances. ➔ Innovative Features: AR lenses, maps, Bitmoji, and Spotlight drive engagement. Impressive, isn't it? Now, have you ever worked on such complex infrastructure? Share your experiences below. #aws #softwareengineering #softwarearchitecture #cloudcomputing

Start building in AWS! I just published a new Pluralsight hands-on lab: “Create and Subscribe to an Amazon SNS Topic” Here's what you'll actually do (not watch, not read… DO): → Design the start of a production-ready SNS notification setup → Configure pub/sub messaging from scratch → Wire up a Lambda function as a subscriber → Validate message delivery end-to-end by publishing a message to your topic You can take the lab two ways: 1. Guided Mode 2. Challenge Mode Both modes offer the ability to check your progress as you complete each objective! You'll be working in a real AWS account with real services, which, in my opinion, is how cloud skills actually stick. I explain a bit more within the video below, and you can get to the link in the comments 👇