Do these problems sound familiar?
- Your company works with a lot of paper documents or scans. Data from these documents need to be entered into a computer system daily.
- Manual data transcription is tedious, time-consuming, and prone to errors.
- Your full-time employees spend too much time moving data, and their work is inefficient.
- Moving data is monotonous and results in high employee turnover. This, in turn, leads to expensive recruitment processes. The newly employed work with lower effectiveness.
Introducing iDoc
iDoc is a solution using AI/ML (Artificial Intelligence and Machine Learning) algorithms to…
- automatically read and categorize document scans
- and fill out databases in company systems with correct information.
How does iDoc work?
Both AI and an employee verify extracted data
Data is sent to other systems
iDoc replaces the process of manually transcribing information from documents into databases
The algorithm reads the information in the document and recognizes (categorizes) the document type. Next, it collects data contained in the document and saves them in iDoc’s database. Then, this database can be integrated with other databases and systems in your company.
iDoc is available on-premise or as a cloud service.
iDoc’s database can complement your existing database or remain a standalone database.
Manual versus automated data entry
During a single working day
A person can categorize about 3100 document pages.
An AI algorithm can categorize over 30 000 pages with at least 96% accuracy.*
A person can read and enter about 2000 attributes into a database.
An AI algorithm can read and enter ten times as many attributes (about 20 000), maintaining approx. 85% accuracy.*
*100% accuracy can be ensured by a person verifying automatically processed data
What kind of information can iDoc read?
iDoc reads documents to recognize all types of information defined for a particular document category. This includes:
Address data (postal addresses, plot numbers, precinct numbers, etc.)
Dates and periods of validity
Other dates in a document
Details about objects in a document
(object name, quantity, type, area, size, etc.)
Data about contracting parties
(company names, owners, representatives, notaries, social security numbers, etc.)
Numbers identifying contracts and others (investment documentation, land and mortgage registries, commissions, etc.)
Information about the contract’s subject (legal status, contract type, termination notice, etc.)
Information about payments (net and gross prices, recurring payments, etc.)
Other detailed data
The range of extracted information can be expanded with more document categories and data types according to your business or industry needs.
How do we ensure the security of your data?
We emphasize the security of our solutions and the protection of our customers’ data. We follow security recommendations and best practices derived from ISO 27001, NIST, and other standards.
In the software development process, solutions are tested according to the OWASP TOP10 and OWASP ASVS methodologies. We use static code analysis, vulnerability verification of images, and of third-party components.
We adhere to a strict separation between development and production environments, to which only solution administrators have access.
Administrative access is protected by two-step verification, with phishing-resistant authentication methods.
We use full separation of each of our client’s data, which prevents unauthorized access by individuals from another organization.
Would you like to know more about extracting information from documents using artificial intelligence? Check out our blog.
Who is iDoc for?
iDoc is a solution for you if…
iDoc will bring desired results in companies that own many paper documents and need the data to be entered into computer systems in order to be further processed and used for daily work.
Examples of such companies:
Insurance companies
Banks
City halls and municipal offices
Property, geodesy, or accounting departments at companies from the following industries:
A solution tailored to your specific needs
iDoc is based on AI/ML algorithms that can be trained depending on your needs. This is why the solution can be adjusted to read different document types specific to various industries or even a particular company.
Examples of application areas
As-built technical documentation
Road occupancy permits
Real estate lease documentation
What can you achieve with iDoc?
5 to 10 times faster data entry from documents into a database
thanks to automation
Facilitated access to up-to-date, detailed information from documents
quick and easy attribute search
More complete databases
easy process of data entry equals more digitized documents
Improved data quality
eliminated human errors
Accelerated processes at your company
due to effortless and faster access to real-time data
Increased control over processes
insights derived from documents help to control schedules, key dates, issued decisions, etc.
Better use of your employee potential
employees can use the time they spent on tedious data entry on more rewarding tasks
Lower recruitment and training costs
lower turnover of employees who work on data entry
Better management of working time
iDoc can process documents outside of working hours (during the night and without taking breaks)
See how iDoc can improve your company operations
iDoc provides convenient and fast access to in-depth, up-to-date information in paper documentation. Use its full potential, improve your company management, and make better strategic decisions.
Would you like to see your documentation on a map?
If your company owns documentation related to geographically dispersed objects, accessing document folders through a map can significantly improve your work.
You can achieve this with LocDoc, another solution using AI/ML for automatically reading and categorizing scanned documents. Based on identified location, it places documents on a map creating a spatial database.
iDoc and LocDoc are a part of the GlobIQ platform for building and maintaining digital network asset models.
Frequently asked questions:
What is the licensing model?
Sales model and the place where the architecture is maintained:
• On-Premises.
• SaaS
An example in a SaaS solution
Price components (per month):
Infrastructure maintenance, administration and technical support:
Dependent on:
• number of licenses,
• SLA level,
• infrastructure
Variable fee for document processing (OCR, Categorization, Reading attributes):
Dependent on:
• number of pages,
• amount of information to reading
Number of pages to be used.
• Specified quantity pages / year
• % discount for buying a package for 1 or 2 years
Can you process cityhall building permit, ID and property deed?
Yes, we can build models to read these types of documents
Does one model cover all requirements in iDOC? Or do we have one model for one requirement?
We train a separate model for each attributeinformation. Same for each document category.
Is the application GUI available in Romanian ?
Currently no. Within a week we can prepare such a GUI. We need another week to isolate the next version of the system and instal it on MV.
Do we train models based on user decisions changing the choice of AI? What does this procedure look like?
iDoc does not have a user interface where it could configure the models itself. This is a complicated activity. To maintain a minimum level of effectiveness, the models are created and trained by Data Science specialists from Globema.
