Commercial data was spread across France/US spreadsheets, statuses and notes. I designed and developed the CRM independently with AI assistance, to connect records, responsibilities and next actions.
GetOut CRMA commercial tool, built for the way we work.
I designed and developed GetOut’s CRM independently with AI assistance: product scope, interface, data model, pipeline integration and tests. It replaces scattered spreadsheets with a usable commercial application.
Explore the case study ↓I framed dense prospect lists, filters, assignment, a pipeline and a detailed record. The product is a custom Next.js/TypeScript application, not a purchased CRM configured for GetOut.
A dedicated Supabase/PostgreSQL base links prospects, commercial owners, contacts, tasks and activity history. Imported statuses are normalised; unknown values remain reviewable instead of being silently reclassified.
The prospecting workflow has its own database. Read-only FDW views expose lead data to the CRM. Suggested prospect–lead matches are confirmed manually and stored on the CRM side.
Follow the prospecting workflow →The lead profile connects identity, location, facilities, contacts and provenance with qualification, sequence status and M1–M4 drafts. Details are loaded when needed rather than treating every record as a ready-to-contact lead.
Draft review and the actions for marking a reply or taking over a lead use dedicated controlled paths. They do not write to the pipeline through FDW. Roles and active profiles control access.
I made the product and integration decisions and used coding agents to implement iteratively. Type checking, interface tests and synthetic SQL access fixtures distinguish intended behaviour from tested behaviour.
The 18 October consolidation reports the CRM and Lead 360 L1 deployed. Advanced approval-safety work remained local; broader synchronisation and autonomous outreach are not claimed. Production has not been re-inspected for this portfolio.
Inspect a commercial record.
Illustrative interface with fictional data, reconstructed for this case study. It is not an authentic CRM screenshot.
Complexe A
- Qualification
- Review required
- Contact
- Fictional sports service / no real email
- Next action
- Resolve the organisation’s identity.
The result and the next step.
CRM and Lead 360 L1 delivery are reported in the 18 October consolidation. Read-only FDW and controlled actions are distinct. Later approval-safety changes retained their own local release status.