Case Study

AI-Powered Extraction with Human Oversight

Tiarna partnered with IndiVillage to transform lease abstraction from a manual bottleneck into a fast, accurate, AI-powered operation.

TIARNA

Tiarna, a Southern California commercial real estate firm managing over 16 million square feet of property, needed to eliminate the administrative burden that was holding back its growth. IndiVillage developed an AI-powered lease abstraction tool that transformed a painstaking manual process into a fast, accurate, and scalable operation - making Tiarna the first company globally in its sector to deploy this capability.

15-20 minutes

per lease, down from hours of manual work

98%+

accuracy across 100+ extracted data points

~$0.50

per lease processing cost

Challenge

Leases Measured in Hours, Not Pages

Commercial lease abstraction is one of property management's most time-consuming tasks. Each lease document contains dozens of critical data points - rent escalations, CAM charges, insurance requirements, parking provisions, security deposits, letters of credit, and special terms such as rent abatements - all of which must be extracted accurately and fed into property management systems like Yardi and MRI. For a portfolio the size of Tiarna's, the cumulative administrative burden was significant.

The manual process consumed hours per lease and required deep familiarity with commercial real estate terminology and document structures. Variance identification in monthly reports added further overhead. With Tiarna targeting 25% revenue growth, the team needed to redirect time from administrative processing toward client relationships, deal-making, and portfolio expansion. The challenge was not just efficiency - it was freeing the operational capacity to grow.

Solution

AI-Powered Extraction with Human Oversight

IndiVillage built a custom AI-powered lease abstraction tool that extracts over 100 data points from each lease document, including insurance coverage limits, parking allocations, rent escalation schedules, CAM charges with base year and carve-out provisions, security deposits, letters of credit with expiration dates, and rent abatement periods. The system processes multiple leases simultaneously and includes natural language search across uploaded documents, enabling users to query their lease portfolio conversationally.

The tool was designed around a human-in-the-loop model: AI extracts and presents data, users verify and correct, and the system learns from each correction to improve over time. Browser automation provides seamless integration with Tiarna's existing Yardi and MRI systems. Iterative development incorporated detailed feedback from Tiarna's operations team - including Scott Alleborn, Eric Pae, and Toan Nguyen - with enhancements for letter of credit handling, CAM base year extraction, and controllable expense exclusions refined through multiple review cycles.

Results

The First in Their Sector, Globally

The tool reduced lease abstraction time from hours to 15-20 minutes per document, with 98%+ accuracy and a processing cost of approximately $0.50 per lease. Tiarna can now upload and process hundreds of historical leases without capacity constraints, with the system handling multiple documents simultaneously. The reasoning feature provides transparent verification, showing users exactly where each data point was extracted and why - building confidence in the output from day one.

For Tiarna's leadership, the impact goes beyond operational efficiency. By eliminating the administrative bottleneck of lease abstraction, the team has reclaimed the capacity to focus on client relationships, portfolio strategy, and the growth targets that manual processing had been constraining. Tiarna is now the first commercial real estate company globally to deploy AI-powered lease abstraction at this level, and discussions are underway for phase two - extending the platform's capabilities deeper into Tiarna's operational workflows.