REST & GraphQL APIs — versioned, documented (OpenAPI/Postman), tested.
service · api
APIs and integrations
that hold up in production.
The API is the product underneath the product. I build the documented, versioned, retry-safe backends that apps and partners actually depend on.
who this is for
Sound familiar?
- Mobile teams needing a proper backend behind the app
- Businesses connecting systems — payments, SMS, CRMs, ERPs, WhatsApp
- Platforms exposing data to partners or the public
- Products with a flaky integration that keeps dropping callbacks
the honest bit
What it costs.
Single-integration work starts around KES 50,000–120,000; full backend APIs with auth, docs and admin tooling scope higher. Existing-integration rescue audits are quoted after a short review — tell me what's misbehaving.
deliverables
What you get.
Payment integrations — Daraja, Stripe, Paystack, PayPal — with idempotent callbacks and reconciliation.
Messaging integrations — WhatsApp Cloud API, SMS gateways, email pipelines.
Third-party connections — maps, logistics, government or industry APIs your workflow needs.
Webhook infrastructure — ingestion endpoints that verify signatures, dedupe retries and never lose an event.
Auth & security — OAuth2, JWT, API keys, rate limiting and audit logging as standard.
how it ships
The process.
step 1
Contract first
Endpoints, payloads, error cases and ownership agreed in an OpenAPI doc before build.
step 2
Build + test harness
Sandboxed integrations with replayable fixtures — your team can test without calling Safaricom.
step 3
Failure-mode hardening
Retries, timeouts, idempotency and dead-letter handling — where most integrations actually live.
step 4
Docs & handover
Postman collection, environment runbook, monitoring hooks. 30-day defect warranty included.
proof
Relevant work.
Not mockups — systems running in production.
// related reading
faq
Straight answers.
Can you fix an existing broken integration?
Yes — integration rescue is common work: callbacks that drop, payments that double-post, tokens expiring mid-request. I audit the failure path first, then fix the cause, not the symptom.
Do you write the docs too?
Yes — every API ships with an OpenAPI spec or Postman collection. Undocumented APIs are how systems die.
What stacks do you build APIs on?
Laravel/PHP, Node.js/Express and Django/DRF — I match your existing stack or recommend one for the team that'll maintain it.
How do you handle Safaricom/Africa-specific quirks?
Token caching, callback signature checks, idempotent writes, retry queues, transaction-status polling — the parts the official docs mention in a footnote and production makes critical.
Describe what needs to talk to what — I'll design the contract and the failure modes.
Written scope, honest pricing, 30-day defect warranty on everything I ship.