找到 187 篇相关文章
AWS has recently introduced CloudFormation express mode, a deployment option that can reduce infrastructure deployment times by marking stack operations complete once resource configuration is applied, rather than waiting for full resource stabilization. By Renato Losio
Recap. Part 1 framed the supervision problem. Part 2 architected the four-model system on AWS and counted the hardware. Part 3 put the gate on the page in fifteen lines of PyTorch. Now the question that decides whether any of it mattered: does the gate actually earn its keep - and what would it cost to know? This is the close. A verification design, the stability instrumentation that separates a real evaluation from a lucky one, and the FinOps reality that is the reason this whole series is a blueprint instead of a benchmark. SDAR makes two claims, not one It's tempting to reduce the paper to a single number. The reported gains over plain GRPO are real and worth stating - their numbers, not mine : roughly +9.4% on ALFWorld, +7.0% on Search-QA, +10.2% on WebShop accuracy. But the headline percentage is the less interesting claim. The one that matters is the second: SDAR avoids the training instability that naive GRPO+OPSD falls into . The whole point of the gate Part 3 was to keep the teacher's noisy rejections from destabilizing training. So a verification that only measures final task success has tested half the paper. You have to measure stability too - and most reproductions don't. That two-pronged claim dictates the experiment. The verification: three arms, not two You cannot prove SDAR's thesis with a before/after. You need three training runs, identical except for the supervision: Arm What it is What it proves A. GRPO Plain agentic RL, no teacher The baseline the +9.4% is measured against B. Naive GRPO+OPSD Teacher distillation, no gate The instability SDAR claims to fix C. SDAR Gated distillation Part 3 Beats A on score and beats B on stability Drop arm B and you can show SDAR beats GRPO, but you've quietly deleted the paper's actual contribution - there's no evidence the gate did anything a plain teacher wouldn't. Arm B is the expensive arm nobody wants to run and the one that makes the result credible. Metrics that actually test the claim Metric What it mea
An error with the cloud computing giant’s billing operation caused some customers’ monthly bills to rise from a few cents to billions of dollars.
Apple filed a trade secrets lawsuit against OpenAI last Friday, and it’s not messing around. The complaint alleges a pattern of misconduct reaching all the way up to OpenAI’s chief hardware officer and claims more than 400 former Apple employees now work at the company. OpenAI’s response so far has been carefully hedged, and the timing couldn’t be worse with the company reportedly eyeing an IPO […]
If someone asked me ten years ago what it takes to break into cloud, I would have said "get certified and hope someone gives you a chance." I was wrong. And I watched dozens of freshers follow that exact advice, collect a certification, then sit in interviews unable to explain why they chose one architecture over another. The problem was never knowledge. It was proof. Proof that you can gather requirements from a confused client, design something that works, and actually build it with your own hands. AWS just launched something that helps close that gap. And two of these credentials cost nothing. Table of Contents What AWS SimuLearn Badges Actually Are Why This Matters More Than Another Certification The 12 Badges Available Right Now The Free Starting Path I Would Follow Today The LinkedIn Advantage Nobody Is Talking About For Career Switchers: Your Existing Skills Are the Cheat Code My Honest Take After 10 Years in Cloud What AWS SimuLearn Badges Actually Are SimuLearn is not another video course. Not another multiple-choice exam. You sit in a simulated client meeting powered by generative AI. A virtual customer explains their business problem. You ask questions, uncover requirements, handle objections, and propose an architecture. The AI evaluates your communication, your technical accuracy, and your decision-making in real time. Then you build the solution. In a live AWS environment. Not a sandbox with three buttons. The real console. After that, an automated validation confirms your solution actually works. Complete every assignment in a learning plan, and AWS issues you a badge through Credly. Automatically. No exam booking. No proctored test. Just demonstrated capability across the full workflow. Each badge represents the entire journey: customer conversations, architecture design, hands-on building, and validated outcomes. Not a single quiz. Not one lab. The whole thing. Why This Matters More Than Another Certification I hold 7 AWS certifications. Let me tell
Some Amazon customers logged on Friday to a surprise bill estimate claiming that they owed the tech and cloud giant billions in fees.
When building Cypress automation that interacts with AWS services, the first step is verifying that your test framework can successfully authenticate and communicate with your AWS account. In this article, you'll learn how to connect Cypress to AWS and perform a simple authentication test using AWS Security Token Service (STS) and the GetCallerIdentity API. This approach helps confirm that: AWS credentials are correctly configured. Cypress can invoke AWS SDK operations through Node.js tasks. The automation environment is connected to the expected AWS account. Establishing this connection first provides a solid foundation before automating interactions with services such as AWS Lambda, Amazon S3, Amazon DynamoDB, Amazon SNS, or Amazon SQS. Prerequisites Before getting started, ensure you have: Node.js installed A Cypress project Valid AWS credentials: AWS Access Key ID AWS Secret Access Key AWS Session Token (if using temporary credentials) AWS Region Note:If you're running Cypress in an AWS environment (such as AWS CodeBuild, an EC2 instance with an IAM role, or GitHub Actions using OpenID Connect), you may not need to provide credentials manually. The AWS SDK can automatically retrieve credentials from the execution environment. Step 1: Install the AWS SDK Install the AWS STS client package: npm install @aws-sdk/client-sts For this connectivity test, we only need the AWS Security Token Service (STS) client. The package provides: STSClient – Creates a client for communicating with AWS STS. GetCallerIdentityCommand – Returns details about the authenticated AWS identity associated with the configured credentials. Step 2: Configure AWS Credentials For local development, create or update your cypress.env.json file. { "AWS_ACCESS_KEY_ID" : "" , "AWS_SECRET_ACCESS_KEY" : "" , "AWS_SESSION_TOKEN" : "" , "AWS_REGION" : "" } Populate the file with your AWS credentials. Example: { "AWS_ACCESS_KEY_ID" : "your-access-key" , "AWS_SECRET_ACCESS_KEY" : "your-secret-key" , "AWS_SES
Amazon Web Services has recently introduced AWS Continuum, a new integrated security platform to automate the discovery, enforcement, and remediation of security issues across codebases, dependencies, and applications. AWS Continuum launches with four agentic capabilities, aiming at the entire vulnerability lifecycle: penetration testing, code review, threat modelling, and code vulnerabilities. By Gianmarco Nalin
From Zero to a Working EKS Pipeline: Terraform, Ansible, and GitLab CI/CD (and Everything That Broke Along the Way) I recently built an end-to-end deployment pipeline on AWS EKS using Terraform for infrastructure, Ansible for configuration, and GitLab CI/CD to tie it all together. On paper, that sentence sounds clean. In practice, it took several rounds of "why is this failing" before it actually worked. This post is not a "here's how EKS works" tutorial. There are plenty of those. This is the version with the failures left in, the quota limits, the IAM permission walls, the pods that wouldn't schedule, and the resources that refused to die. If you're building something similar, I'm hoping this saves you a few hours of confused Googling. Repo: gitlab.com/nenyeonyema/terraform-eks-ansible-cicd What I Was Building The goal was a full IaC-driven pipeline: Terraform to provision the EKS cluster and supporting AWS infrastructure (VPC, node groups, IAM roles) Ansible to handle configuration tasks on top of the provisioned infrastructure GitLab CI/CD to automate the whole thing — plan, apply, configure, deploy — on every push Simple enough in theory. Four separate blockers said otherwise. Blocker #1: Free Tier ASG Restrictions The first wall I hit was with the Auto Scaling Group for my EKS node group. AWS Free Tier limits how much compute you can provision, and my initial node group sizing quietly ran into those limits — the kind of failure that doesn't always throw an obvious, single-line error. Fix: I resized the node group to stay within Free Tier boundaries and got explicit about instance types and desired/min/max capacity in Terraform, instead of leaving Auto Scaling to make assumptions I couldn't afford. Lesson: If you're building on Free Tier, hardcode your capacity expectations early. Don't let the defaults surprise you later. Blocker #2: EKS Private Endpoint Access By default, EKS clusters can be configured with private-only API server endpoint access. That's grea
You've probably registered a domain with Route53 or another popular registrar. You pick something...
MCP for AWS Security Engineers: Build a Read-Only Security Hub Triage Agent For AWS-heavy security work, I would start with AWS Agent Toolkit for AWS and the managed AWS MCP Server , not a custom MCP server. The reason is practical. AWS now provides a managed MCP path that can connect AI coding agents to AWS documentation, AWS APIs, AWS skills, and existing IAM credentials. The Agent Toolkit also provides plugin-based setup for supported agents such as Claude Code and Codex. For security teams, that is the right starting point because the enforcement point remains AWS IAM, not the model. The initial operating model should be strict: Read-only first. No production write authority. No access to secrets. No raw customer PII or sensitive incident logs in prompt context. No automatic remediation. No AI-approved suppression, exception, merge, deploy, or risk acceptance. Human review and CI/CD remain the release authority. That is the same posture I would use for a governed Claude Code or Codex rollout: named identities, SSO, scoped credentials, default deny, tool approval, audit logs, and security evidence tied back to tickets, pull requests, CI logs, and cloud findings. What we are building This article walks through a practical security workflow: A read-only Security Hub triage assistant that helps a junior security engineer produce a daily or weekly findings summary, remediation backlog, and evidence pack without allowing the agent to modify AWS. The agent will be able to: Read AWS Security Hub findings. Group findings by account, severity, product, resource, and control. Explain why a finding matters. Draft remediation tickets. Draft a Slack-ready summary. Produce local markdown, CSV, and JSON evidence files. The agent will not be able to: Suppress findings. Archive findings. Mark findings resolved. Disable Security Hub standards. Modify IAM, S3, EC2, KMS, GuardDuty, Inspector, or Config. Deploy remediation. Run destructive scripts. Approve risk acceptance. This is no
I've been building AI agents for a while now, and streaming responses to a UI has always been the painful part. In previous projects I tried API Gateway streaming, Lambda response streaming, and even AppSync Events via an agent tool call to notify the UI. I also looked at adding my own WebSocket API through API Gateway, which requires managing $connect , $disconnect , and $default routes, storing connection IDs in DynamoDB, and posting messages back through @connections . All of these approaches felt like too much ceremony for what should be simple. That's when I found that AgentCore has built-in WebSocket support. The browser can just connect directly to the agent. No middleman. I built WearCast to demonstrate this functionality. This is an AI agent that helps you pick out what to where based on the weather. I've found this very helpfu when packing for a trip. The code for this application is public and you can find the full implementation on this GitHub repo . The problem with the traditional approach We usually build app applications as we do REST APIs User → REST API → Lambda → AI Service → Lambda → REST API → User Every message makes a round trip through multiple intermediaries. The response waits until the entire generation is complete, and then it all comes back at once. For a chat interface, this feels sluggish. Users stare at a spinner while the model generates hundreds of tokens they could already be reading. Even if you add Server-Sent Events or long-polling, you're still stitching together a real-time experience on top of infrastructure that seemed like overkill. I wanted something better. The architecture Here's what I ended up with: ┌─────────────┐ ┌──────────────────┐ │ React UI │─── JWT ─▶│ API Gateway + │──▶ Lambda (presigned URL) │ (Cognito) │ │ Cognito Auth │ └──────┬──────┘ └──────────────────┘ │ │ WebSocket (SigV4 presigned URL) │ ← No middleman! Direct connection → ▼ ┌──────────────────────────────────┐ ┌────────────────────┐ │ AgentCore Runtim
AWS and Anthropic have released the Claude apps gateway for AWS, a self-hosted control plane that centralizes identity, policy, telemetry, routing, and spend caps for Claude Code and Claude Desktop. The gateway runs as a single stateless container and routes inference to Amazon Bedrock or Claude Platform on AWS. By Steef-Jan Wiggers
The Problem Real411 is a South African platform where citizens report digital harms: misinformation, incitement, hate speech, and harassment. When someone submits a complaint, it needs to be reviewed by multiple people, assessed against legal criteria, and resolved with a public verdict. The process must be transparent, auditable, and fair. I joined this project early and worked on it extensively over a long period. A senior solutions architect consulted on the database schema design. There was a cloud person who helped with parts of the infrastructure. Other coworkers contributed at different stages. I spent most of my time on the API layer and the frontend components. This article covers the architecture decisions I worked with, what I learned from the senior architect's design choices, and how the system evolved. The Status Machine Most applications model status as a column on a table. You update the value and the old state is gone. That works for simple workflows but fails when you need to know not just where a complaint is now, but how it got there and who made each decision. The senior architect who consulted on the database design suggested an append only status log. Instead of a single status column, the complaint_status table records every transition as a separate row. Each row has the status code, the user who made the change, a timestamp, and optional notes. The current status is derived by querying the most recent row. I implemented this pattern across the API layer. Every status transition became an insert operation rather than an update. It took some adjustment to shift from mutable state to event sourced state, but the benefits were immediate. Auditing became straightforward. The state machine also became easier to implement because each transition is a simple insert with a business logic check, not a conditional update. The schema has seventeen status codes covering the full lifecycle: received, claimed, under assessment, pending secretariat review,
The Problem Sports scoring sounds simple. One team scores a point, the number goes up, everyone sees it. But when you build it as a web application that needs to work on courtside tablets, spectator phones, and wall mounted displays simultaneously, with voice commands and tap controls, the architecture becomes more interesting. The project was Scoring AI, a voice enabled match scoring application for sports courts. Players start a match, share a link, and control the scoreboard from any device. The backend handles real time state synchronization, optimistic locking, idempotent score updates, rate limiting, and WebSocket broadcasting. The team was small. Me and a coworker who handled the CI/CD side. We were at the same level, both full stack, and we designed the system together. He focused on the deployment pipeline and infrastructure automation. I focused on the application layer, the real time system, and the frontend. But the architecture decisions were shared. This article covers the technical decisions we made and how patterns from previous projects influenced them. Why DynamoDB for Live Matches The match scoring data is different from the business data around it. A match lasts about an hour, gets updated frequently, and needs to be read by many viewers at once. After the match is complete, it is archived and rarely accessed. I had seen what happens when you put high frequency state updates into a relational database on a previous project. Row locks, contention, connection pool exhaustion. For Scoring AI, we used DynamoDB for the live match state and PostgreSQL for everything else. The hot path needed fast writes, optimistic locking, and automatic cleanup of abandoned matches. DynamoDB provides all of these. The version field on each match record acts as an optimistic lock. Every score update is a conditional write that checks the version has not changed. The cold path uses PostgreSQL through Kysely for user profiles, subscriptions, pricing plans, payment histor
The Next Chapter When I left the waiter job and returned to engineering, I knew I wanted something different. Not just a different job, but a different way of working. The kind where your location does not limit the problems you can solve. I found that in Thailand, working for a South African company called Exonic. Why Bangkok After Dubai, I wanted somewhere with a lower cost of living where I could build runway while working remotely. Bangkok checks that box. The city is a hub for remote engineers. The internet is fast. The infrastructure works. The street food is better than any restaurant I have ever worked in. I arrived with a laptop and a clear goal: find a remote role where I could work on meaningful projects without being tied to a physical office. Landing the Role at Exonic Exonic is a technology consulting company based in South Africa. They serve clients across multiple industries and geographies. When I found the opening, it matched exactly what I was looking for: full time remote, exposure to diverse projects, and the chance to work across the full stack. The interview process was practical. System design discussions, technical assessments focused on AWS and modern frontend frameworks, and conversations about how I approach end to end delivery. I got the offer and accepted it immediately. As a full time remote employee, I was embedded in Exonic's engineering team. My day to day involved building cloud native solutions for their clients, designing architectures on AWS, and shipping production systems across the entire stack. The team was distributed, and the work required communicating clearly across time zones. Three Continents Through One Company Exonic's client base spans the globe. Over my time there, I built production systems touching three different continents. One project was Scoring AI , a voice enabled match scoring application for sports courts. Players start a match, share a link, and control the scoreboard using voice commands. I worked on th
The Hidden Cost of Manual IAM Review Most teams don't track how long they spend reviewing IAM policies. When I started measuring it on my own team, the numbers were worse than I expected. A thorough manual review of one IAM policy takes 10 to 15 minutes. Not a quick scan. A real review: read every statement, trace every cross-account trust, verify every condition key, check for privilege escalation paths, confirm the resource ARNs match what you think they should. At 4 engineers touching IAM once a week, that's 4 hours a month. 48 hours a year of senior engineers reading JSON documents. And that's the optimistic case. Add a security incident. Add an audit. Add the emergency Friday-afternoon policy change that needs review before deploy. The real number is higher. What manual review misses The problem isn't just the time. It's that humans are bad at repetitive structured-data review, especially under time pressure. Here are the things I've seen slip through manual IAM reviews on production systems: iam:PassRole with no condition. This is the big one. PassRole lets a principal pass a role to a service — and if there's no iam:PassedToService condition, that role can be passed to any service that accepts roles. Including services the attacker controls. The reviewer saw the action, mentally categorized it as "role stuff," and moved on. It was statement 47 of 52 — the reviewer had already been reading policies for 40 minutes. Wildcard resource with sensitive actions. s3:* on Resource: "*" is obvious. s3:GetObject on "arn:aws:s3:::*-backup/*" with a wildcard in the bucket name — that's subtle. The reviewer reads it as "restricted to backup buckets" and moves on. But the wildcard means any bucket ending in -backup , including ones in other accounts if cross-account access is configured. Missing aws:SourceArn on Lambda invocation permissions. When you grant another service permission to invoke your Lambda function, you need aws:SourceArn to prevent the confused deputy
Hachette, Cengage, Elsevier, and other publishers allege that Google trained its AI on copyrighted works without the necessary permissions.
You know the drill for adding an MCP server to a project: dig the exact command string out of the docs, hand-write a .mcp.json with an absolute path you'll typo once, restart the editor, and discover no tools showed up because the server expected a config file you haven't created yet. Plenty of MCP servers lose their would-be users somewhere inside that loop. Infrawise collapses the whole loop into one command. It's an open-source tool ( npm ) that statically analyzes your codebase, AWS infrastructure, and database schemas, then exposes that context to AI coding assistants over MCP — so Claude Code knows your actual partition keys, GSIs, and indexes instead of guessing from source files. This post is about the part that usually kills tools like this before they deliver any value: setup. Section 1: One command, four steps npm install -g infrawise # or skip install and use npx cd your-project infrawise start --claude start does four things, in order: 1. Probes your environment. If there's no infrawise.yaml in the project, it generates one. It reads AWS_PROFILE if set; otherwise it looks at your configured AWS profiles — one profile means zero questions, several means one prompt asking which to use. That's the entire interview. (If you want the full guided wizard instead, infrawise start --interactive runs it.) 2. Runs the analysis. It scans your AWS services, database schemas, and codebase, builds a graph of services, tables, indexes, and query patterns, and runs rule-based analyzers over it. No LLM is involved in this step — extraction and analysis are deterministic, so the same infrastructure always produces the same graph. 3. Writes .mcp.json to your project root. This is the file you'd otherwise write by hand: { "mcpServers" : { "infrawise" : { "command" : "infrawise" , "args" : [ "serve" , "--stdio" , "--config" , "/absolute/path/to/infrawise.yaml" ] } } } 4. Opens Claude Code. Claude Code reads .mcp.json automatically and starts the session with all 21 infrawise
This essay was written with Nathan E. Sanders, and originally appeared in The Guardian . Opposition to AI data centers has emerged as a primary theme in US politics, one that—surprisingly—doesn’t fall along party lines. We applaud people coming together for constructive debate on any issue, and agree that communities need to evaluate whether any economic benefits these data centers bring is worth their costs. Still, we worry that a focus on data centers obscures the larger impacts of AI on people’s lives: the concentration of power of AI companies, and their widespread political and financial influence...