We connect your systems. Then we make the connection hard to break.
Bridge is integration middleware. We connect Zendesk, Airtable, CRMs and the other systems your work actually runs through — built once for your case, then either handed to you, licensed to you on a dedicated German instance, or operated by us.
Most integrations fail quietly. A webhook is missed, a retry loops, two systems overwrite each other, and nobody finds out until a customer does. Bridge is built around that failure mode rather than around the happy path — which is why most of what follows is about what happens when something goes wrong.
How it works
Three steps, and the honest version of each
No jargon wall. This is the whole architecture, in the order an event travels through it.
Events in
Your systems send us a webhook the moment something changes. Every delivery is signature-verified before it is accepted, and stored in a durable queue — so a restart, a deploy or an outage on our side loses nothing.
Signature verification · durable Postgres queue · duplicate deliveries are a no-op
Transform and route
A declarative mapping decides what travels and what it becomes. Rules decide whether an event belongs in the other system at all. An event that matches no rule is not silently synced and not silently dropped — it stops and asks a person.
Field mapping · conditional routing · loop prevention · per-record ordering
Verified writes out
The write is made against the record we have already matched — never a guessed one. A failure that waiting can fix is retried with backoff; a failure that waiting cannot fix goes straight to the dead-letter queue where you can see it, rather than being hidden by an hour-long retry ladder.
Matched-record writes · retry with backoff · permanent failures surfaced immediately
What “managed” means
Six things, built in
“Managed” is a word that usually means nothing. Here is exactly what the Bridge does. On the licence and operated tiers it runs on your dedicated instance; on the one-time tier we build the same behaviours onto your own platform and you run them.
EU hosting
Your instance runs on a dedicated server in Germany (Hetzner, Nuremberg or Falkenstein). Data residency is a property of where it runs, not a policy we promise.
Retry and backoff
Transient failures — a rate limit, a timeout, a five-hundred — retry on an exponential schedule. Failures that retrying cannot fix are separated out and raised, not buried.
Duplicate protection
A repeated webhook delivery is a no-op. A change we made ourselves is recognised and not echoed back, so the two systems cannot talk each other into a loop.
Error logging
Every event carries its own timeline: received, attempted, failed, retried, written. You can trace one record end to end and read the reason in plain language.
Operational monitoring
Queue depth, dead-letter depth, throughput, success rate and latency — with honest denominators: a rate over zero events reads as “—”, never as 0%.
Updates and maintenance
We keep the instance patched, the dependencies current and the connection working as the systems either side of it change their APIs.
Pricing
Four tiers, published
Two things are priced separately: the implementation — built once, for your systems — and what happens after. Own it yourself, licence it, or have us operate it.
In EUR, excluding VAT. Published on purpose: you should not have to sit through a call to find out whether this is the right order of magnitude.
Build on Your Stack
We build it properly on your platform — you own everything after.
€4,000one-time
Nothing recurring
- Platform
- Your existing automation platform; the tool licence is yours
- What we build
- The full integration, both directions — field mapping, conditional routing, duplicate protection, idempotency, retry and backoff. The parts DIY implementations get wrong.
- Handover
- Documentation and a guided handover
- Hosting & operation
- Yours — your platform licence, your hosting, your 2am
- Changes after handover
- Billed at €150/h; modifications you make fall outside warranty
Bridge Licence
The Bridge, deployed for your systems and licensed to you.
€2,000/ year
+ €4,000 one-time setup
- Deployment
- The Bridge running for your system pair on a dedicated German instance
- Setup covers
- Field mapping, conditional routing rules, transformation logic, and end-to-end validation against your real systems
- The annual licence covers
- Hosting, updates and security patches
- Volume
- Unlimited syncs, unlimited fields, unlimited mappings within the pair
- Support
- Best-effort via the support portal. No response-time guarantee.
- Changes & incidents
- Mapping changes and incident work billed at €150/h
- What you licence
- The right to run it — not the source code
Bridge Licence Pro
The licence with room to move — and the option to run it yourself.
€3,500/ year
+ €4,000 one-time setup
- Includes
- Everything in Bridge Licence
- System pairs
- Up to 3
- Where it runs
- Our dedicated German instance, or an on-premise deployment on your own infrastructure — still licensed to run, not the source code
- Included changes
- 5 hours of consulting or mapping changes per year, then €150/h
- Support
- Priority in the support queue — still best-effort, with no response-time guarantee
- What this is not
- No monitoring and no guaranteed response. That is Bridge Operated.
Bridge Operated
Everything in the licence, plus somebody accountable for it.
€1,200/ month
+ €4,000 one-time setup
- Includes
- Everything in Bridge Licence
- Monitoring
- Alerting on failure rate and queue growth
- Support response
- Same business day
- Mapping revisions
- Included
- System pairs
- Up to 3
- Who it is for
- Clients who want someone accountable around the clock — priced accordingly
On volume and rate limits.Sync volumes are unlimited on our side. Throughput is ultimately governed by the API rate limits of the systems being connected — these are set by each vendor and vary by your subscription tier with them. Bridge queues and paces requests within those limits automatically; if your volume outgrows what your vendors’ plans allow, that is a plan conversation with those vendors, not an integration failure.
On the boundary. On the one-time tier, operational responsibility transfers at handover and the boundary is stated in writing before work begins. On the licence tiers, support is best-effort and change work beyond what the tier includes is billed hourly — also stated in writing.
Support, plainly
What the licence includes, and what it does not
- Both licence tiers include updates and security patches, for as long as they run.
- Support is best-effort through the support portal. There is no guaranteed response time on either licence tier — Pro is prioritised in the queue, which is not the same as a guarantee.
- Mapping changes, new rules and incident investigation are billed at €150/h. Licence Pro includes five hours a year before that starts.
- If you need a guaranteed response and revisions included as a matter of course, that is the Operated tier.
The licence price is possible because nobody is on call for it. If you need someone on call, that is the Operated tier, priced accordingly.
Isolation & compliance
Your instance is yours alone
Every client gets a dedicated Bridge instance: its own server, its own database, its own credentials. There is no shared application tier, no shared queue and no shared database — not as a configuration setting, but because a client’s deployment is a separate machine. Both the server and the database sit in Germany, so EU data residency is a fact about where the software runs rather than a clause in a document. If you leave, we tear the instance down and the data goes with it; if you would rather hold the database yourself, that is a supported arrangement and we will say so in the contract.
Tell us what needs connecting
Name the two systems and what has to travel between them. We will tell you whether it is a Bridge job, and what we would not be able to do — before you commit to anything.
admin@kwayzer.com · Built in Berlin · Hosted in the EU