Service 06 — API & integration

Nothing in your
business should
live on an island.

We integrate your data and resources into your private and public sites and applications — designing APIs worth consuming, and connecting the systems your team already depends on.

What is included

  • Contract first
  • Authentication that fits
  • Resilience
  • Webhooks in both directions
  • Data mapping and ETL
  • Observability

The integration is where projects go wrong

Feature work is predictable. Integration work is where estimates die — because the other system's documentation is out of date, its rate limits are undocumented, and its idea of a customer record does not match yours.

We treat every integration as its own small project with its own risk. Before we quote it we read the API, make real calls against a sandbox, and find out where the surprises are. That is why our integration estimates hold when other people's do not.

Designing an API worth consuming

If you are publishing the API rather than consuming it, the audience is developers and the product is the contract. Versioning, pagination, error shapes, idempotency and rate limits are the interface. We design those first, document them properly, and treat a breaking change as the serious event it is.

Deliverables

What you
actually get.

01

Contract first

An agreed schema — OpenAPI or GraphQL — before implementation, so the client and server teams can build in parallel.

02

Authentication that fits

API keys, OAuth or signed requests, scoped to least privilege and rotatable without downtime.

03

Resilience

Retries with backoff, idempotency keys, circuit breakers and dead-letter queues, because the other system will go down.

04

Webhooks in both directions

Verified signatures, replay protection and a delivery log, so nobody has to ask whether an event arrived.

05

Data mapping and ETL

Reconciling two systems that disagree about what a customer is, with the conflicts surfaced rather than silently resolved.

06

Observability

Per-integration dashboards and alerts, so a partner's outage is visible to you before it is visible to your customers.

Shapes of the work

Where this
usually lands.

Integration typeCommon examplesThe usual trap
PaymentsStripe, PayPal, ACH providersWebhook ordering and duplicate events
ERP / accountingNetSuite, SAP, QuickBooks, XeroRate limits and nightly batch windows
CRM & marketingSalesforce, HubSpot, KlaviyoField mapping drift after somebody edits a form
LogisticsCarriers, 3PL, warehouse systemsAddress normalisation and partial shipments
Internal legacyA twenty-year-old database or a fixed-width fileNo sandbox, no documentation, one person who remembers

Stack

What we build
this with.

We work in your stack when you have one. These are our defaults when the choice is ours.

RESTGraphQLOpenAPIWebhooksOAuth 2.0Node.jsLaravelPythonRedisRabbitMQ

Questions

Before you
commit.

The other system has no API. Now what?
Then we look at exports, database access, SFTP drops or, as a last resort, a carefully-scoped scraping layer with monitoring. We will tell you plainly how fragile each option is before you choose.
How do you handle rate limits?
Queueing, backoff and batching, with the limit written into the design rather than discovered in production. We also alert when you approach it, not when you hit it.
Can you document an API we already have?
Yes. We read the implementation, produce an OpenAPI or GraphQL schema from reality, and flag the places where the behaviour and the intent disagree.
What if the partner changes their API?
Version pinning, contract tests in CI and alerting on shape changes. You find out from us, not from a customer.

Ignition

Need api & integration?

Tell us the problem, the users and the systems already in place. We will identify the right discovery or build step.