Technology worth knowing today.
Artificial Intelligence
Google Announces Second Edition of Site Reliability Engineering Book Updated for AI and Cloud Complexity
What happened. Google has released a completely rewritten second edition of its Site Reliability Engineering (SRE) book to mark the ten‑year anniversary of the original publication. The new edition incorporates a decade of lessons learned inside Google and across the industry, offering fresh guidance for operating today’s complex production environments that include AI workloads, microservices, multi‑cloud, serverless and platform‑engineering practices. Google says the shift in challenge has moved from adopting cloud infrastructure to maintaining high reliability amid growing system complexity and interdependencies.
Why it matters. The updated SRE book provides operators with current best practices for keeping AI‑intensive services reliable, which is critical as more organizations embed machine learning models into production pipelines. By addressing modern complexities such as deep dependencies and rapid change, the guidance helps teams avoid costly outages and maintain performance while scaling AI workloads across hybrid and multi‑cloud infrastructures. This supports safer, more predictable AI adoption in enterprise settings.
Who should care. Developers · Cloud architects · Platform engineers · Data engineers · Technology leaders · People learning AI
Source: Google Cloud
Cloud Computing
AWS Announces Lambda Web Functions for HTTP‑Native Serverless Workloads
What happened. Amazon Web Services announced AWS Lambda Web Functions, a new Lambda form factor that lets developers run standard HTTP web applications built with Express, Hono, Fastify, Next.js or any Node.js framework without rewriting code into event handlers. Users can deploy with a single command, and Lambda provides automatic scaling, multi‑region distribution, and active CPU billing. The launch adds to the existing Lambda Event Functions and Lambda MicroVMs, giving three distinct programming models on the same Firecracker isolation technology.
Why it matters. This development means developers can move existing web apps and API backends to a serverless platform without the usual refactoring or adapter work, preserving their current code and framework choices. By handling scaling, multi‑region placement and billing based on actual CPU use, Lambda Web Functions reduce operational overhead and cost uncertainty for variable‑traffic workloads. It expands the range of applications suited for Lambda beyond event‑driven functions, offering a familiar HTTP model while retaining the isolation and management benefits of the Lambda service.
Who should care. Developers · Cloud architects · Platform engineers
Source: AWS
Data Agent Kit Brings Agentic Data Science to Google Cloud IDEs
What happened. The Data Agent Kit, introduced as a plugin for coding agents and a Visual Studio Code extension, equips data scientists with MCP tools that connect agents to Google Cloud services and Skills markdown files that encode best‑practice instructions. Using the kit inside an IDE, users can ingest raw transaction logs from Google Cloud Storage, load them into BigQuery with Spark, train fraud‑detection models, score new records, and schedule the pipeline to run daily—all without leaving the editor.
Why it matters. By bundling data‑service connectors, reusable skill templates, and IDE‑native tools for querying, model training, and pipeline deployment, the Data Agent Kit reduces the context‑switching friction that typically slows collaborative data‑science work on Google Cloud. Teams can build, test, and productionize end‑to‑end analytics workflows faster, keeping code, data, and orchestration steps in a single environment.
Who should care. Data engineers · Developers · Cloud architects · Technology leaders
Source: Google Cloud
DevOps
Cloudflare Makes Logpush Transformers Generally Available for In‑Pipeline Log Shaping
What happened. Cloudflare announced that its Logpush log export service now includes generally available Transformers, letting users run SQL queries to filter, redact, and reshape logs before they are sent to SIEMs, analytics platforms, or object storage. The feature, rolled out during the company’s recent Birthday Week, moves log‑processing logic into the export pipeline so teams no longer need a separate service to perform those tasks.
Why it matters. By embedding transformation logic directly in Logpush, Cloudflare removes the need for teams to deploy, scale, monitor, and maintain a separate processing layer. This reduces operational overhead and eliminates the cost of exporting, ingesting, and storing log records that would later be discarded. Users can now tailor log shape and sensitivity controls at export time, which cuts downstream data volume and simplifies compliance workflows.
Who should care. Developers · Cloud architects · Platform engineers · Security teams · Administrators · Data engineers
Source: Cloudflare
Running AI Agents as Kubernetes Resources with kagent
What happened. kagent, a CNCF Sandbox project, lets developers define AI agents as Kubernetes custom resources. By writing an Agent manifest with a system prompt, model, and tool list, applying it creates a pod that speaks the Agent-to-Agent protocol. The post demonstrates an agent that can generate another agent’s YAML, showing that the process is repeatable without new frameworks. Security notes mention the default unsecured auth mode and shared PostgreSQL password.
Why it matters. Treating agents as Kubernetes resources brings AI workloads into familiar DevOps practices: version‑controlled manifests, automated rollouts, and native observability. This lowers the barrier for teams to experiment with agent‑based workflows while keeping deployment, scaling, and rollback consistent with other cloud‑native services. It also highlights where security considerations must be added before production use.
Who should care. Developers · Platform engineers · Cloud architects · Technology leaders · People learning AI
Source: CNCF
Today’s takeaway
Today’s stories show the industry folding new AI-driven workflows into established operational patterns rather than inventing new ones from scratch. Google’s rewritten SRE book, a decade after the original, says the core discipline hasn’t changed even as AI workloads raise the complexity of what’s being kept reliable. AWS’s new Lambda form factor and Google’s Data Agent Kit both reduce friction by meeting developers where they already work — standard HTTP frameworks and the IDE — rather than asking them to adopt new models. Cloudflare’s Logpush Transformers do the same for log pipelines, moving a task teams used to bolt on separately into the export step itself. The kagent story is the most direct version of this pattern: it defines AI agents as ordinary Kubernetes custom resources, version-controlled and deployed like anything else, though its own security notes are a reminder that convenient defaults still need hardening before production use.