The problem
An heir-hunting firm — probate research — depended on manually researching estates to judge which cases were worth pursuing. Slow, expert-dependent work that capped throughput at about 70 leads a week regardless of effort.
Every lead the team did not have time to read was a case that simply never got evaluated.
What we built
- 1An Airtable-based probability-scoring system that evaluates and ranks each estate lead.
- 2Legal API integration pulling the records the scoring depends on.
The results
- Capacity: 70 → 250 leads per week, a 4x increase
- 500+ hours a year of manual research eliminated
- 87% scoring accuracy measured against expert judgment
- 4 additional deals closed from cases that would never have been reviewed
Why it worked
The firm's edge was its judgment about which estates pay off. Encoding that judgment into a scoring model meant it could be applied to every lead instead of only the ones the team had hours to read. The 87% accuracy figure is the important one — it is what made the team willing to trust the ranking.