How a Big Four accounting firm
reduced lease abstraction time by 80%
Customer
Major audit and management consulting company
Goal
Automate the extraction of information from lease agreements with a fast, reliable solution that also frees up human resources to work on higher value tasks
Challenges
- Due to both international and US regulation changes (the new IFRS 16 and ASC 842 lease accounting standards), the firm has the task of helping its customers to put all leases on the balance sheet
- Large enterprises might have hundreds of thousands of lease agreements, which must be scoured for relevant information
- Lease agreements are not standardized: different lessors use multiple different formats
- Manual processing of very high volumes of complex lease agreements is costly and prone to human error
- Keyword-based search systems are not well suited to the lease abstraction task because the same type of information can be expressed using many different terms
Solution
Incorporation of the SemanticPro Extract & Analyze Engine into the existing contract-processing workflow, forming the lease abstraction component of that workflow.
The engine automatically:
- Processes documents in English, German, and Spanish
- Handles PDF files from scanned lease agreements that have undergone optical character recognition (OCR)
- Extracts precise, key data, such as amounts, dates, and addresses
- Extracts defined types of clauses and classifies agreements accordingly
- Infers missing information in agreements from the presence of other pieces of data
- Feeds all extracted information into the firm’s lease-management tool, enabling that tool to:
- Instantly fill a summary sheet with data for each agreement
- Display each summary sheet beside the corresponding agreement, for quick review by a subject matter expert
- Highlight the extracted information in the displayed agreement
Results
- 80% reduction in lease abstraction time
- Significant reduction in lease abstraction costs
- Reassignment of expensive human resources to tasks of higher added value
- Quick provision of accurate data from high volumes of leases, assisting subject matter experts in preparation of clients’ balance sheets
The Cortical.io Difference
- The solution was seamlessly integrated into the accounting firm’s contract-processing workflow
- Extraction models could be trained very quickly, based on annotations by the firm’s subject matter experts on only a handful of sample lease agreements
- The solution instantly extracts quality structured data from lease agreements
- The subject matter experts have control over improving the lease abstraction process by verifying, correcting, and adding extraction targets
