David Okafor

Meridian was losing 15% of clients annually but had no early warning system. Account managers only learned about issues when clients threatened to leave or didn't renew.

Reduce client churn from 15% to 10.3% saving $1.8M in annual recurring revenue

Built a health score model in Looker that combines engagement metrics (meeting attendance, response times, NPS), project metrics (timeline adherence, budget usage), and relationship signals (stakeholder changes, complaint frequency). Claude generates weekly account risk summaries and suggests proactive interventions.

Client churn dropped from 15% to 10.3% in the first year. Early intervention saved 7 accounts worth $1.8M in annual revenue.

Meridian's regulatory compliance required quarterly reports to 3 bodies, each with different formats and data requirements. Two full-time analysts spent most of their time on data gathering and report formatting.

Created n8n pipelines that aggregate compliance data from HR, finance, and project systems into Airtable. Claude transforms raw data into each regulator's required format and generates the narrative sections. Automated submission reminders and audit trail logging.

Compliance reporting went from requiring 2 dedicated FTEs to 10 hours/quarter of review. Both analysts were reallocated to higher-value strategic work. Zero compliance findings in the next 2 audit cycles.

Leadership spent 3 hours every Monday in a status update meeting where department heads read metrics off slides. Half the room was checked out and the metrics were always a week stale.

Connected all operational data sources to Looker via n8n. Claude generates a daily written brief highlighting metric changes, anomalies, and recommended actions. Department heads get auto-tagged on metrics in their domain. Async comments replace synchronous discussion.

Eliminated the weekly all-hands (recovering 120+ person-hours/month) and leadership reported making faster decisions with daily data vs. weekly snapshots.

Project margins varied wildly (from -5% to 45%) but no one could explain why. Finance saw the numbers after the fact and project managers had no real-time visibility into budget burn.

Built an Airtable model that tracks time entries, expenses, and change orders against original project budgets in real time. Claude analyzes burn rate patterns and flags projects heading for overrun before they hit the danger zone. n8n syncs data from the time tracking system hourly.

Identified $800K in untracked scope creep across active projects. Average project margin improved from 22% to 31% in 2 quarters by catching overruns early.