Skip to content
Zeeshan Akhtar

API & Business Integrations

Custom API integrations for what off the shelf automation can't cover

n8n, Make and Zapier cover most cases. Some don't have a prebuilt connector, need to turn unstructured documents into usable data, or need a small purpose-built application instead of another automation subscription — that's where custom API work comes in.

Is this the right fit?

This is the right fit when two systems need to talk and no existing connector does it, when documents or scraped content need to become structured data, or when the actual need is a small custom tool rather than a workflow.

Problems this solves

  • Two systems need to talk to each other and no prebuilt connector exists
  • Documents or unstructured content need to become structured, usable data
  • The business needs a small custom application, not another automation platform subscription
  • Appointment and finance data live in separate systems with no automatic way to keep them aligned

What I build

  • A Cliniko to Google Sheets to Notion sync built around a healthcare practice's real appointment and billing workflow, where no single connector covered the case
  • A Python scraper written directly against a source with no connector — handling pagination, inconsistent page structure, retries and rate limit backoff — feeding structured lead data into an n8n workflow that dedupes it and delivers it where the team already works (how that Python and n8n split works)
  • A small purpose built application when the real need is a tool rather than another workflow — FreelancerOS, the system I built and run for my own work, is the clearest example
  • REST API and webhook integration work where a platform's own connector doesn't cover the case

Tools

APIsWebhooksPythonPostgreSQLCustom applications

How this works

Every engagement starts with how the process actually runs today, then moves through design, build, and testing against real data before handover. See the full five step process for the details.

Frequently asked

What counts as a 'custom' integration versus a normal automation?

If n8n, Make or Zapier already has a reliable connector for both systems, that's normal automation work. Custom integration means writing directly against a platform's API or webhooks — usually because no connector exists, or because the case needs logic a pre-built connector can't express.

Can you build a small internal tool instead of another automation?

Yes — sometimes the right answer genuinely is a small purpose built application rather than another automation platform subscription. FreelancerOS is one I built for my own work, and that kind of tool is part of what a custom integration engagement can cover.

Do you work with healthcare or other regulated data?

The Cliniko healthcare workflow automation project connects appointment and billing data across systems the practice already uses, moving data that's already legitimately accessible rather than introducing new access.

What if the system I need to connect has no API or connector at all?

Then the integration is written against whatever the source actually exposes. The lead generation system is the clearest example: a Python scraper handles pagination, inconsistent page structure, retries and rate limit backoff, and hands structured records to an n8n workflow that dedupes and delivers them. Where a clean connector exists I use it; custom code is for the cases where it doesn't.

Will my team be able to maintain a custom integration without you?

Every engagement ends with documentation and a handover so routine changes don't depend on me. That's the same standard as the ongoing Habfract automation, where each workflow is documented so the team can operate it independently.

Have a process like this to automate?

Tell me what's slow or manual right now. I'll tell you what it takes to fix it.

Discuss a Project