MaxOps AI

The problem

The beta platform had critical Softr ↔ Airtable sync issues — Kanban pipeline reverting, lead-status inconsistencies, and data overwrites — plus unreliable automations and an architecture that couldn't scale beyond a single account.

It wasn't production-ready.

What we built

We stabilized the system and re-architected it into a scalable, multi-tenant SaaS platform for brokerages.

Agents, brokerage admins (tenants), and a platform admin each see only their authorized data on a shared infrastructure.Fixed Softr ↔ Airtable synchronization: corrected field mappings and write-back logic, and separated automation-controlled fields from manual edits to stop overwrites and Kanban reversion.

• Fixed Softr ↔ Airtable synchronization: corrected field mappings and write-back logic, and separated automation-controlled fields from manual edits to stop overwrites and Kanban reversion.
• Multi-tenant architecture letting many brokerages and agents share one infrastructure with role-based data isolation.
• Secure OAuth email for agents (Gmail & Outlook) — send/receive without sharing credentials.
• Calendly scheduling with webhook-triggered pipeline updates.
• Round-robin lead assignment for even distribution across agents.
• Lead-generation pipelines via scraping tools and Typeform feeding structured Airtable data, plus a documented future-state architecture.

How we built it

We restructured the Airtable relationships and Softr filters so each tenant sees only their own data, implemented role-based access control, and corrected the sync logic causing overwrites.

OAuth integrations connected agent inboxes securely; Calendly webhooks updated the CRM pipeline automatically; and Zapier / Make handled reliable triggers, Typeform-to-Airtable mapping, and round-robin assignment. A full system-architecture document mapped data flow from capture through nurture, scheduling, and conversion.