The client is a Texas-based land and real estate verification company. It delivers robust and accurate title records data through a technology-enabled platform to title companies, attorneys, investors realtors and financial institutions. It provides insights into important details of real estate and mineral rights transactions lying with county courthouses.
To maintain a comprehensive, accurate and close to real-time updated database of land records and leasing information, the technology company aggregated millions of data points like deed, deed of trust, assignments, etc. from title documents resting with different courthouses in the USA.
The scope of work was very high in terms of:
- Capturing hundreds and thousands of data points from unstructured and fragmented information sources
- Validating them for accuracy
- Ensuring they were as recent as possible and
- Standardizing and cleansing the data
They partnered with Hitech to put in place a seamless and agile on-going land records data aggregation process to populate its database.
- Segregating and tagging crucial title information for a huge range of title documents across categories required proper understanding of courthouses terminology and domain expertise.
- Zoning in on the required information from an unstructured or hand-written 100+ page document for every single record made the entire data capture process complex and time-consuming.
- Maintaining high level of data extraction and entry accuracy since the slightest mistake made public could invite legal penalties.
Initially, Hitech data specialists implemented a structured workflow for multi-sourced manual data capture from county courthouses documents, cleansing, validation and upload in excel spreadsheets. We developed an automated quality check validation tool to spot and rectify formatting, data validation and compliance errors in the spreadsheets.
The client then transitioned to an in-house application. They were so happy with the performance of the custom validation tool that they incorporated it into their application.
Hitech continues to populate cleansed and validated property data in the new in-house application. Detailed weekly reports are also shared to provide granular visibility across key production metrics.
Hiring and Training
- A team of 20 trained and skilled data specialists was put in place for processing hundreds of thousands title documents. To meet the voluminous demand, the team worked in multiple shifts to best leverage time-zone benefits.
- They were imparted detailed domain knowledge on required data fields, complex property related terms, state and county regulations, etc.
- Scanned files in TIFF and PDF formats of title documents were shared by the client as input documents. The Hitech team accessed the information over client’s remote application using secured credentials.
- Manually capture/enter data from title documents such as Deed, Deed of Trust, Assignments, etc. in pre-defined templates.
- Cleanse and standardize the data to maintain consistency and prepare it for use as a saleable database.
- Conduct a double-layered manual as well as rule-based validation and verification process by quality controllers to ensure data authenticity against the source documents.
- Processed up to 1.5M+ title documents which translated to approx. 8M+ records till date with 99.5% accuracy.
- Share detailed reports on number of documents scraped vs. received, accuracy and productivity with the client on weekly basis.
Tools and Technology used- Windows Virtual Desktop and Customer Data entry application
8M+ records updated till now
99.5% database accuracy achieved
Cost optimized due to high efficiency and seamless processes
Data accuracy enabled compliance conformity