Back to Case Studies
Cybersecurity & AITechnical Build

AI-Powered Development of a Multi-Client Compliance Tracking Platform

A full-stack compliance dashboard built with AI-assisted development in ~1 month — replacing manual cross-referencing with real-time, automated progress tracking across multiple clients and 9 compliance phases each.

Next.js 15Supabasen8n AutomationNotion APIAI-Assisted Dev
~1 mo
Concept to Production
3
Active Clients
9
Phases Tracked
Zero
Manual Sync
Our Process
01 Strategize
Requirements & Architecture
02 Build
AI-Assisted Development
03 Deploy
Multi-Client Launch
04 Scale
Self-Service Onboarding

01The Client

Secure Cloud Innovations (SCI)

A cybersecurity consulting firm managing SOC 2 compliance engagements for multiple clients simultaneously — including Cofertility, Geofund, Disputed, and future clients.

02The Challenge

SCI needed a way to track compliance project progress across multiple clients, each with 9 distinct phases containing dozens of tasks managed in Notion. The existing workflow was entirely manual.

What Was Broken:

  • Manual cross-referencing. Consultants had to dig through Notion databases, calculate phase timelines, and manually communicate progress to stakeholders.
  • No unified view. There was no single place to see compliance progress across all clients.
  • No real-time tracking. Stakeholders had to wait for manual status reports.
  • No automated data sync. Every update required manual effort — no system connected Notion to any client-facing interface.

03What We Built

Full-Stack Web Application

  • Next.js 15 (React 19) frontend with Tailwind CSS
  • Supabase (PostgreSQL) backend with edge functions
  • Dynamic routing (/[clientSlug]) supporting unlimited clients from a single codebase
  • Expandable phase cards with progress bars and kanban-style task boards

Real-Time Data Pipeline

Notion → n8n → Supabase → Dashboard

  • Custom n8n workflows pulling task data from 9 Notion databases per client
  • Supabase Edge Function that ingests, transforms, and stores task data
  • One-click Refresh button triggering webhooks to re-sync data on demand
  • PostgreSQL views calculating phase status dynamically (Not Started / On Track / Behind / Completed)

Multi-Client Architecture

  • Three active clients each with unique Notion databases and n8n workflows
  • New Client configuration form with real-time Notion URL validation
  • Automated n8n workflow deployment via API — new clients get a fully configured 9-phase sync workflow without manual setup

Authentication & Tooling

  • Role-based password protection: admin, per-client, and user-level access
  • HttpOnly cookie sessions with 30-day persistence and server-side auth checks
  • 30+ diagnostic scripts for database verification and data migration

04Technical Complexity Solved

  • Phase timeline calculation: Deriving phase start/end dates from individual task days_from_start values relative to a project start date.
  • Cross-system data integrity: Handling Notion ID formats, UUID conversion, missing fields, and database constraint violations across three different client workflows.
  • Status logic edge cases: Correcting for phases with incomplete Notion data using client-side fallbacks.

05Tech Stack

Frontend
Next.js 15, React 19, Tailwind CSS, Lucide
Backend
Supabase (PostgreSQL), Edge Functions (Deno)
Automation
n8n (workflow orchestration), Webhooks
Data Source
Notion API (9 databases per client)
Auth
Cookie-based with HttpOnly, role-based access
Deployment
Vercel, Supabase Cloud, n8n Cloud

06AI's Role in Development

Claude served as the primary development partner throughout — architecting the system, writing application code, building database schemas, debugging complex data pipeline issues (Notion field mapping, n8n transform syntax errors, UUID conversion), designing the UI, and iterating on edge cases in real time.

The human-AI collaboration was conversational and iterative: developer described what was needed, Claude built it, they tested together, and Claude debugged and refined until each piece worked end-to-end. Entire platform designed, built, and iterated across 19 commits in ~1 month — a pace that would typically require a small engineering team over several months.

07Results & Impact

~1 Month
Initial commit to multi-client production
3 Clients
Onboarded with hundreds of tasks tracked
Self-Service
New clients added via UI with automation
Real-Time
Stakeholders check progress anytime

Zero ongoing manual sync — data flows automatically from Notion (where consultants work) to the dashboard (where clients check progress), eliminating the need for manual status reports entirely.

Ready to build something like this?

No decks. No fluff. Just shipped systems.