Java, architecture, DevOps & the craft of building — read, distilled, and written for you.
Java / JVM10 pieces
Your Loom App Quietly Became a Thread Pool Again: A Field Guide to Virtual Thread Pinning
Virtual thread pinning causing silent throughput ceiling — critical operational gotcha for any Loom migration in production.
tutorialFoojayRead the brief →
New candidate JEP: 401: Value Objects (Preview)
JEP 401 Value Objects Preview is a major Valhalla milestone directly impacting JVM performance and design.
explainerr/javaRead the brief →
Identifying JDK Value Class Candidates
Directly tracks JDK value class adoption path — critical for performance-oriented Java engineers watching Project Valhalla.
tutorialInside Java (Oracle)Read the brief →
How We Beat HotSpot Performance (By Cheating, But Not Like That)
ParparVM achieving HotSpot parity via frameless C codegen — deep JVM performance engineering with real benchmark data.
tutorialFoojayRead the brief →
Identifying JDK value class candidates
Identifying JDK value class candidates is directly useful as Valhalla/JEP 401 approaches adoption.
explainerr/javaRead the brief →
Java News Roundup: TornadoVM 5, JHipster, Google ADK, OmniFish Build of Payara, Introducing Vidocq
TornadoVM 5 GA, GraalVM updates, Keycloak — actionable Java ecosystem releases to evaluate.
explainerInfoQ ArchitectureRead the brief →
Spring Office Hours Podcast: S5E18 - The Latest from OpenAI, Anthropic and Spring AI 2.0
Spring AI 2.0 updates alongside OpenAI/Anthropic changes — directly relevant for Java AI integration work.
explainerSpring BlogRead the brief →
Reachability Annotations for generating GraalVM metadata
GraalVM native-image metadata generation at compile time directly eases native Spring Boot deployments.
tutorialr/javaRead the brief →
What's new in Plinth 0.17.0?
Plinth 0.17 covers migration safety with Flyway/Mongock and architecture boundaries — directly applicable.
explainerr/javaRead the brief →
The Java Story | The Official Documentary
Official Java documentary — valuable historical and cultural context for a Java senior engineer.
explainerLobstersRead the brief →
Architecture & Patterns7 pieces
Temporal Is to Your Code What a Database Is to Your Data
Temporal as durable execution layer for workflows — strong architectural analogy, highly relevant for distributed systems design.
explainerFoojayRead the brief →
Aurora DSQL: Scalable, Multi-Region OLTP
Aurora DSQL multi-region OLTP is directly relevant to distributed systems architecture and AWS cloud design.
explainerLobstersRead the brief →
🛑⚡ When NOT TO USE Event-Driven Architecture (EDA)
Honest trade-off analysis of EDA — useful for architecture decisions, avoids hype, covers observability/consistency costs.
explainerFoojayRead the brief →
DSLs Enable Reliable Use of LLMs
DSLs as guardrails for LLM code generation — useful pattern for reliable AI-assisted systems.
explainerMartin FowlerRead the brief →
Article: Comprehension at AI Speed: Building a Context Store for Evolutionary Architecture
Context stores for evolutionary architecture under AI speed — addresses real systemic risk he faces.
explainerInfoQ ArchitectureRead the brief →
How DoorDash Built an AI Shopping Assistant That Doesn’t Rely on the LLM Alone
DoorDash's multi-agent MCP architecture for production AI assistant — solid reference implementation.
explainerInfoQ ArchitectureRead the brief →
Why every AI agent decision needs a receipt
Audit trails and observability for AI agent decisions maps directly to distributed systems reliability concerns he cares about.
explainerThe New StackRead the brief →
DevOps / Cloud / SRE19 pieces
AI Agents with Cloud Credentials Are Outrunning Billing Guardrails Built for Human-Speed Mistakes
AI agents with cloud credentials causing $14k AWS bills — critical SRE/security risk to address now.
explainerInfoQ ArchitectureRead the brief →
AI Agents with Cloud Credentials Are Outrunning Billing Guardrails Built for Human-Speed Mistakes
$14K AWS bill from stolen keys + autonomous agent cloud spend — critical security/cost risk for any AI infra he runs.
explainerInfoQ AI/MLRead the brief →
Warm Up Fast, Run Lean: Vertical Scaling for Java on Kubernetes with Azul Prime and Kedify
Vertical pod autoscaling for Java on Kubernetes with JVM-aware warm-up — directly addresses production scaling gaps.
tutorialFoojayRead the brief →
AWS Ships Claude Apps Gateway as Self-Hosted Control Plane for Claude Code and Claude Desktop
Self-hosted control plane for Claude Code with IAM, spend caps, telemetry — directly useful for platform teams deploying agentic tools.
explainerInfoQ ArchitectureRead the brief →
Article: Removing a Hidden Round Trip from a Multi-Region AWS API
Eliminating hidden round-trips in multi-region AWS APIs for failover — high-value reliability engineering insight.
tutorialInfoQ ArchitectureRead the brief →
Building a Custom Metrics Exporter for Kubernetes
Custom K8s metrics exporter for queue-based scaling — directly actionable for platform engineering work.
tutorialKubernetes BlogRead the brief →
Running a self-hosted LLM in Kubernetes with vLLM
Self-hosted LLM on K8s with vLLM — highly practical for running private models in his platform.
tutorialCNCFRead the brief →
From intent to enforcement: Lessons from operating Kubernetes controllers at scale
Deep operational lessons on Kubernetes controllers at scale — directly actionable for platform engineering work.
explainerThe New StackRead the brief →
You don’t have a deployment problem. You have a validation problem.
Validation vs deployment framing is a sharp architectural insight directly applicable to platform engineering and CI/CD.
explainerThe New StackRead the brief →
Kubernetes won the container decade. Google’s Agent Substrate wants the next one.
Google's Agent Substrate as next-gen Kubernetes for agents is a potentially significant platform shift to track early.
explainerThe New StackRead the brief →
Presentation: From OTEL to SLMs: Distilling Frontier Model Behaviour from Production Telemetry
OTEL + SLMs distilling production telemetry into LSPs — directly bridges observability and AI tooling for his interests.
explainerInfoQ AI/MLRead the brief →
AWS Continuum to Enable Agentic Code Security for Enterprises
AWS Continuum automates security across codebases/dependencies — relevant for platform/SRE work.
explainerInfoQ ArchitectureRead the brief →
HAMi becomes a CNCF incubating project
HAMi CNCF incubation — GPU virtualization/sharing for K8s AI infra, directly solves real cost problem.
explainerCNCFRead the brief →
Is a Pod the right deployment unit for an AI agent?
Pod-per-agent K8s deployment model — concrete architectural decision for running agents in production.
explainerCNCFRead the brief →
Operating OpenTelemetry at scale with OpAMP
OpAMP for managing OTel Collectors at scale — directly addresses observability scaling pain.
explainerCNCFRead the brief →
Why smarter AI caching sometimes makes everything slower
AI caching performance pitfalls are real operational concerns for anyone running LLM infrastructure in production.
explainerThe New StackRead the brief →
AWS will now watch Microsoft’s cloud for you
AWS Security Hub now monitors Azure resources — useful for multi-cloud security posture management.
explainerThe New StackRead the brief →
Cloud Native Infrastructure Emerges as the Foundation for Trustworthy Agentic AI
CNCF framing cloud-native infra as the foundation for agentic AI aligns his platform engineering and AI interests.
explainerInfoQ AI/MLRead the brief →
AWS: Inaccurate Estimated Billing Data – $1.7 billion
AWS billing anomaly at $1.7B scale is a real cloud cost/reliability war story worth studying for SRE awareness.
explainerHacker News (250+ pts)Read the brief →
AI / Agents / Skills8 pieces
How DoorDash Built an AI Shopping Assistant That Doesn’t Rely on the LLM Alone
DoorDash's production LLM+agents+MCP architecture is a concrete blueprint he can apply or sell as a service.
explainerInfoQ AI/MLRead the brief →
LLM Tool Call Reasoning Using Embabel Agentic AI Framework
Embabel agent framework for Java with tool-call observability — practical agentic AI in the JVM ecosystem.
explainerBaeldungRead the brief →
OpenAI’s GPT-Red automates prompt injection testing to harden AI agents
Automated prompt injection testing for agents is a concrete security/reliability tool worth evaluating for production agents.
explainerThe New StackRead the brief →
[AINews] Kimi K3 2.8T-A50B: the largest open model ever released; Opus 4.8-class at Sonnet 5 pricing
Kimi K3 largest open model ever, Opus 4-class quality at Sonnet pricing — strong cost/capability signal for LLM tooling choices.
explainerLatent SpaceRead the brief →
What building Shippy taught us about building agents
Practical agent-building lessons from production — directly useful for his AI product work.
explainerHugging Face BlogRead the brief →
Model Routing Is Simple. Until It Isn’t.
Model routing complexity at scale is a real infra problem he'd face building AI-powered services.
explainerHugging Face BlogRead the brief →
QCon AI Boston: Production AI Moves Beyond Prompts to Platforms, Harnesses, and Evals
Production AI ops beyond prompts — evals, context management, harnesses are exactly what he needs building agents.
explainerInfoQ AI/MLRead the brief →
Linus Torvalds on LLM usage in kernel development
Torvalds' take on LLM use in kernel dev — authoritative signal on AI coding tools' real-world limits.
explainerLobstersRead the brief →
AI Dev Tooling9 pieces
AWS Ships Claude Apps Gateway as Self-Hosted Control Plane for Claude Code and Claude Desktop
Self-hosted Claude Code control plane with policy, telemetry, spend caps — exactly the enterprise AI tooling governance he'd need.
explainerInfoQ AI/MLRead the brief →
The cost of saying yes has changed
Critical framework for deciding what AI-generated code to accept; directly impacts his shipping decisions and tech debt.
explainerGitHub Blog (Copilot)Read the brief →
Show HN: Clawk – Give coding agents a disposable Linux VM, not your laptop
Disposable VM sandbox for coding agents directly addresses safe agentic coding workflows he uses.
explainerHN Show HN (100+)Read the brief →
I Asked GitHub Copilot to Profile a Java App. It Found a Bug in My Heap Sizing, and Offered to Fix It
GitHub Copilot used for JVM heap profiling and bug detection — practical AI dev tooling applied to Java perf work.
explainerFoojayRead the brief →
1Password’s new browser integration for Claude changes how AI uses your credentials
1Password+Claude credential integration for agents — solves real auth problem in agentic workflows.
explainerThe New StackRead the brief →
5 Trends That Defined AI Engineering at World’s Fair 2026
AI engineering trends from major conference — agent-centric system design directly applicable to builder work.
explainerLatent SpaceRead the brief →
[AINews] Codex usage up 10x in 6 months to 7M users, +1M in the past ~day; did Codex overtake Claude Code??
Codex vs Claude Code usage comparison — directly relevant for choosing/evaluating AI dev tooling.
explainerLatent SpaceRead the brief →
Better tools made Copilot code review worse. Here’s how we actually improved it.
Copilot code review internals and how tool design shapes agent quality is actionable for his dev workflow.
explainerGitHub Blog (Copilot)Read the brief →
Show HN: Mindwalk – Replay coding-agent sessions on a 3D map of your codebase
3D codebase map for replaying agent sessions is a novel observability tool for agentic coding workflows.
explainerHN Show HN (100+)Read the brief →