Back to Projects
B2B SaaS · Sales

From Airtable duct-tape to a scalable AI sales platform

We diagnosed the lead-gen bottleneck, stabilized the existing setup so revenue never stopped, then rebuilt it as an AI-first product — roughly tripling booked meetings along the way.

For: B2B SaaS sales teams running prospecting on duct-taped toolsDelivered by Z-Flow + a curated specialist team — working MVP in 8 weeks, full product in 12.

Industry

B2B SaaS · Sales

Timeline

8-week MVP · 12-week product

Key Metric

~3× booked meetings

Status

Live

The Challenge

The Challenge

WeProspectify was running its entire prospecting motion on a fragile stack of Airtable, spreadsheets and a few n8n flows. It worked — barely — but it couldn't scale, broke under load, and ate hours of manual work every week. The hard part: the duct-taped system was also the live revenue engine. It couldn't be switched off to be fixed.

  • 20+ hours a week lost to manual lead research and list-building
  • A brittle Airtable + n8n setup that broke as volume grew
  • Outreach scattered across tools with no single source of truth
  • No clean schema to build on — every new channel meant more glue
The Solution

The Solution

We started with a diagnosis, not a rebuild. First we stabilized the existing Airtable + n8n engine so the team kept booking meetings while we worked. Then — underneath the running system — we re-architected it into a scalable, AI-first product on a proper schema, and migrated over without downtime.

Technical Approach

The production rebuild runs on Next.js + Supabase with a clean, scalable schema. n8n orchestrates lead enrichment and multi-channel sequencing across Smartlead, HeyReach and Airtable. OpenAI powers personalization, and an MCP layer lets the system extend and scale with AI agents rather than more manual glue. Stripe handles SaaS billing. Same partner from the scrappy MVP to the production platform.

Results

Results

Booked meetings / month

Baseline~3× baseline
~3× more meetings

Manual prospecting time

20+ hrs/weekLargely automated
20+ hrs/week saved

Time to product

Fragile MVPScalable platform
MVP in 8 wks · product in 12

Key Takeaways

  • Stabilize before you rebuild — the revenue engine kept running the whole time
  • A clean schema and an MCP layer let the product scale with AI agents, not more manual glue
  • One senior partner across MVP and production rebuild meant zero context loss

Tech Stack

Next.jsSupabasen8nOpenAISmartleadHeyReachAirtableStripeMCP

Your project could be next

Let's discuss how we can help you achieve similar results.

From Airtable duct-tape to a scalable AI sales platform