MOUSAServices
Build, repair, and keep it running.
Consultancy, implementation, and automation — the same three buckets the larger partners sell — delivered by the person who will open the Zap.
Zapier
Zaps that stay readable when the team inherits them
Multi-step Zaps, Paths, filters, formatters, webhooks, and Tables. I keep task count honest so you are not paying Zapier for work Make would do cheaper — and I do not force a migration if Zapier is the right home.
| Typical work | Lead capture, CRM upsert, Slack alerts, Stripe events, form → sheet, RSVP sync |
|---|---|
| Good when | The company already lives in Zapier and the flow is mostly linear |
| Watch-outs I fix | Silent filters, Copilot leftovers, duplicate contacts, dead auth |
Make.com
Scenarios for branching, volume, and the visual map
Routers, aggregators, iterators, error handlers, and HTTP when no native module exists. Make is usually the better seat when you can see the whole graph and you care about operations cost.
| Typical work | Shopify/GHL, Stripe/Airtable/Xero, ClickUp onboarding, Etsy ops, enrichment |
|---|---|
| Good when | The flow splits, loops, or needs cheaper volume than Zapier tasks |
| Watch-outs I fix | Empty bundles, inconsistent variables, rate limits, missing error routes |
Rescue
When it “should work” and does not
A large share of engagements start as a broken Zap or scenario. I reproduce the failure on a real record, patch the mapping, and leave notes so the next person is not guessing.
Empty outputs
Make modules that run and return []. Usually an upstream filter, iterator, or mapping — not the module you were staring at.
Duplicates & misses
CRM records created twice, or not at all. Search-before-create and idempotent keys.
Copilot Zaps
Generated flows that look complete and fail on cancel, reschedule, or the second booking of the day.
AI in the workflow
Models inside Zapier and Make — not a separate science project
Classify tickets, draft outreach from your own copy, enrich contacts, generate PDFs, or send audio via ElevenLabs. A person stays in the loop on anything that sends or charges.
Already shipped
Airtable + ChatGPT PDF generation. Claude → audio with Make and ElevenLabs. Custom GPT connected to Clio via Make. n8n contract flows with Sheets and DocuSign.
CRM & revenue ops
HubSpot, GoHighLevel, Airtable, Stripe, Xero, QuickBooks
The unglamorous work that pays for itself: contacts that match companies, price IDs that match invoices, wallet top-ups that become the right tax invoice.
Engagements
How we can work
Fixed-scope build
One workflow, your real data, a defined done. Best for a first automation or a rescue.
Monthly retainer
Monitoring, small changes, and a short path when production breaks.
As needed
When the system is still being discovered and a fixed quote would be fiction.
Next
Bring the stack
Tool names, a sample record (screenshot is fine), and what “done” looks like. I come back with a plan and a timeline.