Pipedrive integration: hire and sales system sync with spend roll-ups
We built a read-only, one-way sync that keeps 4,207 customers from a UK equipment-hire business's on-premise hire and sales system up to date in Pipedrive every 30 minutes. A nightly pass rolls 16,214 invoices into 12-month, 24-month and lifetime spend fields. It never duplicates or deletes: two runs in a row change nothing.
By ExcenityAI · Published
The challenge
Sales had no customer, contact, site or spend data in the CRM. The source system had no usable reporting and no clean master record.
How it works
- 01Read only from the sourceA built-in guard blocks anything except read requests to the client's system.
- 02Map the recordsCustomers become organizations, contacts become people, and sites and finance become fields.
- 03Find, never duplicateMatched by account number, then customer ID, before anything is created.
- 04Update what changedOnly changed fields are written. A watermark in Pipedrive tracks progress, so no database is needed.
- 05Nightly reconcileA nightly pass rechecks everything and recomputes 12-month, 24-month and lifetime spend.
- 06Alerts and retriesFailures alert by webhook and retry on the next cycle.
What we delivered
4,207 customers synced
Plus 16,214 invoices since 2023 rolled up into spend fields for reporting.
Idempotent
Acceptance test: two runs in a row create 0 and update 0 records.
Never deletes
Inactive customers get a label instead, and the CRM client has no delete call at all.
Supervised rollout
Dry-run, single-customer test mode and drift checks before the full load.
Hardened hosting
A hardened server with encrypted credentials and restricted access.
Around 190 tests
All run against fake APIs, never the live systems.
Stack
- Pipedrive API
- REST API (on-prem)
- Node.js
- TypeScript
- zod
- DigitalOcean
- systemd
Related service: CRM & integrations
Related guide: How to sync data into Pipedrive without duplicates
