About HookCastle

A small webhook delivery service, run by a small team in Milan. Started in 2023, public since 2024, paying customers since early 2025.

Why this exists

I'm Claudio. Before HookCastle I worked as a backend engineer at three different companies, and at all three I ended up writing or debugging the same code: a Postgres-backed retry queue for outgoing webhooks, with cron picking up failed deliveries and re-attempting them. It was always 80% there and 20% bugs. The 20% included things like: events delivered out of order to the same endpoint, duplicate deliveries when the worker crashed mid-flight, the queue table never being vacuumed, and an entire afternoon spent figuring out why a customer wasn't receiving notifications (the answer: a 4-month-old Postgres connection pool exhaustion).

The third time I ended up writing this code, in 2023, I extracted it into a separate service for that company's internal use. When I left, I rebuilt it from scratch, made it multi-tenant, and put it on the internet. That's HookCastle.

How it's run

HookCastle is operated from Italy as a small independent project. As of April 2026 we're a small team — myself full-time, plus two part-time collaborators on backend and dashboard. We don't have investors, a sales team, an SDR, or a Slack channel called #growth. The roadmap is decided by what customers actually email about, plus my own opinion when there are no good emails to follow.

Infrastructure-wise, the production service runs on a single beefy VPS in Sofia, Bulgaria, with encrypted off-site backups in a separate EU region. We're EU-hosted on purpose: it's where most of our customers are, it makes the GDPR conversation a non-conversation, and the latency from our region to the rest of the EU is consistently <30ms. We don't use Cloudflare or any US CDN in front of the production service — TLS terminates on our box, certificates are Let's Encrypt, end of stack.

What we work on, and what we don't

We work on:

• Making delivery boring — predictable latency, no surprises.
• The dashboard: searching events, replaying things, debugging customer-side issues.
• SDKs and the CLI.
• Honest pricing and not being annoying about it.

We are explicitly not working on:

• Inbound webhook ingestion (i.e. we deliver, we don't receive on your behalf). There are good products for that — we like Hookdeck.
• Event streaming / Kafka-style pub/sub. Wrong tool, wrong shape.
• A "platform" or marketplace. We sell one thing.
• Per-region data residency outside the EU. Not on the roadmap.

Contact

The fastest way to get a real answer is email: hello@hookcastle.com. It comes to me directly. For technical questions about the API, the docs are usually faster, but emailing is fine too.

For security disclosures: security@hookcastle.com with PGP fingerprint 4F2A 9D1C 8E63 7B5A … (full key in .well-known/security.txt).