The Message That Broke FIFO
FIFO can preserve delivery order and still produce the wrong business state after DLQ replay. How lanes, revisions and recovery policies help.
Engineering · 10 min read · 2026-10-20
Articles
Engineering notes, field guides and perspectives on reliable event delivery.
Engineering notes, field guides and perspectives on reliable event delivery, organized around event delivery, AI and agents, IoT and edge systems, integrations, and comparisons.
Start with a featured guide, browse by topic, or compare Queuey with webhook platforms, cloud queues and building delivery infrastructure yourself.
Featured
- The Guide to the IoT Galaxy — A practical guide to reliability between the device, the edge and the cloud — offline operation, event custody, delayed telemetry and where MQTT's guarantees stop.
- The Agent Worked. The System Failed. — An AI agent can make the right decision and still create the wrong outcome. As agents reach production, AI reliability becomes a distributed systems problem.
- Login used to be application code. For some reason, event delivery still is. — Event delivery often starts with an HTTP request and quietly grows into retries, queues, idempotency, DLQs and recovery logic. Why is this still application code?
- Queuey vs the Alternatives: Choosing Event Delivery Infrastructure in 2026 — Svix, Hookdeck, QStash, EventBridge, Temporal, Kafka or build it yourself? Compare the architectures — and choose the layer you actually need.
Browse by topic
- Event Delivery — Retries, failure handling, replay and reliable delivery infrastructure.
- AI & Agents — Coding agents, agentic systems and reliable AI-generated integrations.
- IoT & Edge — Devices, telemetry, offline delivery and durable buffering at the edge.
- Integrations — SaaS integrations, application architecture and integration plumbing.
- Comparisons — Queuey compared with webhook platforms, cloud queues and internal delivery systems.
Latest articles
We May End Up Owning a Lot More Software.
Coding agents are making first versions dramatically cheaper to build. That may also mean companies end up owning far more internal software than they ever consciously decided to.
Engineering · 4 min read · 2026-10-15
Reliability Wants the Payload. Privacy Wants It Deleted.
Replay needs payload retention. Data minimization pushes the other way. Here’s how to separate delivery history, payload retention and access without giving up reliability.
Engineering · 4 min read · 2026-10-06
How to remove thousands of lines of delivery code with your coding agent
Use your coding agent and queuey advise to find where Queuey fits, implement the first flow and identify delivery code you may no longer need.
Engineering · 3 min read · 2026-09-29
Login used to be application code. For some reason, event delivery still is.
Event delivery often starts with an HTTP request and quietly grows into retries, queues, idempotency, DLQs and recovery logic. Why is this still application code?
Engineering · 5 min read · 2026-09-23
Sanity reduced webhook retries. The reason is more interesting than it sounds.
Sanity shortened its webhook retry window. The interesting question isn't how long to retry, but where durability should live when delivery fails.
Engineering · 4 min read · 2026-09-21
The Guide to the IoT Galaxy
A practical guide to reliability between the device, the edge and the cloud — offline operation, event custody, delayed telemetry and where MQTT's guarantees stop.
Engineering · 9 min read · 2026-09-13
The Agent Worked. The System Failed.
An AI agent can make the right decision and still create the wrong outcome. As agents reach production, AI reliability becomes a distributed systems problem.
Engineering · 5 min read · 2026-09-06
GitHub’s Retry Storm Shows What Retry Logic Is Missing
During GitHub’s August 17 outage, a latent retry bug amplified traffic by approximately 10×. The interesting part isn’t that retries failed. It’s that the system kept retrying after retries had become part of the problem.
Engineering · 4 min read · 2026-08-24
Queuey vs the Alternatives: Choosing Event Delivery Infrastructure in 2026
Svix, Hookdeck, QStash, EventBridge, Temporal, Kafka or build it yourself? Compare the architectures — and choose the layer you actually need.
Comparison guide · 12 min read · 2026-08-19
When your AI tells you to use someone else
Lovable helped us find its own hosting limitation. We moved publishing to Cloudflare — and crawler traffic jumped roughly 100x in two days.
Engineering · 5 min read · 2026-08-17
The Early Partner Advantage
A new service category is taking shape. The advantage is getting there early.
Perspective · 5 min read · 2026-08-09
The Repeatability Gap
Every integration project quietly solves the same problems. The firms that stop rebuilding the same operational foundation will define the next category of integration consulting.
Perspective · 6 min read · 2026-07-24
Your Retry Logic Is Lying to You
A webhook that returns 401, 422, 429 or 503 is not asking for the same treatment. Stop pretending it is.
Engineering · 5 min read · 2026-07-07
From the API Economy to What?
APIs created access. Events create dependency. The next economy will run on reliable signals.
Thought Leadership · 4 min read · 2026-07-03
Web3 was about ownership. The Agentic Internet is about action.
Why the Agentic Internet is fundamentally different from Web3—and why the next generation of software will be designed for agents first, humans second.
Thought Leadership · 3 min read · 2026-06-04
Why HMAC integrations break — and how Queuey makes them reliable
HMAC isn't hard. Integrations are. Here's why signature mismatches happen — and how to remove the entire class of bugs.
Engineering · 5 min read · 2026-04-27
Why database backups don't fix integration failures
Most integration failures are not data problems. They are delivery problems. Event replay fixes what backups can't.
Engineering · 4 min read · 2026-04-13
MCP: Giving AI agents access to real-time events
The Model Context Protocol lets AI agents query live event streams. Here's how it works with Queuey.
Product · 5 min read · 2026-03-04
Event-driven partner integrations at scale
How to onboard partners in hours instead of months using a unified event delivery layer.
Architecture · 7 min read · 2026-02-21
Dead-letter queues: Your safety net for failed events
What happens when event delivery fails after all retries? DLQs keep failed events recoverable instead of silently dropped.
Engineering · 5 min read · 2026-02-09
Compare Queuey
See how Queuey compares with webhook platforms, cloud queues and building delivery infrastructure yourself.
Queuey vs the Alternatives: Choosing Event Delivery Infrastructure in 2026
- Queuey vs Svix: Webhook Infrastructure vs the Async Queue for Your API — Compare Queuey and Svix for webhook delivery, retries, observability, failure handling and pricing. See where each platform is the better fit.
- Queuey vs Hookdeck: Event Gateway vs the Async Queue for Your API — Queuey vs Hookdeck for webhook delivery, queueing, retries, observability, circuit breaking and failure recovery. An honest 2026 comparison.
- Queuey vs Kafka: Event Streaming vs Event Delivery — Queuey and Kafka solve different parts of event-driven architecture. Compare streaming, HTTP delivery, retries, DLQs and external destinations.