Contract management · Production

Generative AI + IDP Extracts Data from Lease Agreements with 82% Accuracy

The problem

A fast-food franchise employed 25 full-time outsourced workers to manually read and re-enter data from 30,000 lease documents annually, a process that was error-prone and required rework, while also needing to meet FASB compliance obligations requiring extraction of more than 350 fields from lease-related documents weekly.

First attempt

The franchise had previously attempted solutions with multiple advisory firms and tested five different technology combinations before finding one that yielded a superior result.

Workflow diagram · grounded in source
1
Lease documents submitted
trigger
“commercial leasing agents submitting lease documents of varying quality, types (such as PDFs, images, and email attachments), and languages. The documents also contain handwriting, checkboxes, and tables, making them difficult to process…”
2
IDP classification and segmentation
ai_action
“ABBYY Vantage is a powerful intelligent document processing (IDP) solution for categorizing the different agreement types and extracting segments of the agreement (Preamble, Premises, Renewal Options)”
3
GenAI field-level extraction
ai_action
“GPT-4 Turbo provided accurate field-level extraction, especially when fed a specific segment of the agreement”
4
Confidence threshold routing
routing
“For the small subset of fields falling outside the configured confidence threshold, ABBYY Vantage's manual review station is utilized”
5
Human review and correction
human_review
“enable human users to correct machine-extracted values”
6
ML model learns from corrections
feedback_loop
“The machine learning model learns from the human-corrected values and improves over time, resulting in reduced manual review requirements”
7
Data entered into lease system
integration
“extract more than 350 fields from lease-related documents weekly and input the data into their lease management system”
Reported outcome

The hybrid IDP and GenAI solution achieved an 82% extraction accuracy rate across 350+ fields and 30K lease documents processed annually, with a human review loop enabling the ML model to improve over time and reduce manual review requirements.

Reported metrics
Extraction accuracy rate82%
Fields extracted350+
Lease documents processed annually30K
Outsourced employees doing manual data entry (prior state)25
Show all 5 reported metrics
extraction accuracy rate82%
fields extracted350+
lease documents processed annually30K
outsourced employees doing manual data entry (prior state)25
manual review requirementsreduced manual review requirements
Reported stack
ABBYY VantageGPT-4 TurboSS&C Blue PrismRPA
Source
https://www.abbyy.com/customer-stories/generative-ai-idp-extracts-data-from-lease-agreements-with-82-accuracy/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

The hybrid IDP and GenAI solution achieved an 82% extraction accuracy rate across 350+ fields and 30K lease documents processed annually, with a human review loop enabling the ML model to improve over time and reduce…

What tools did this team use?

ABBYY Vantage, GPT-4 Turbo, SS&C Blue Prism, RPA.

What results were reported?

Extraction accuracy rate: 82%; Fields extracted: 350+; Lease documents processed annually: 30K; Outsourced employees doing manual data entry (prior state): 25 (source-reported, not independently verified).

What failed first in this deployment?

The franchise had previously attempted solutions with multiple advisory firms and tested five different technology combinations before finding one that yielded a superior result.

How is this contract management AI workflow structured?

Lease documents submitted → IDP classification and segmentation → GenAI field-level extraction → Confidence threshold routing → Human review and correction → ML model learns from corrections → Data entered into lease system.

Related contract management cases