Textkernel’s extremely accurate resume parsing software, Extract!, is constantly being fine-tuned and improved by our large team of research engineers. Read more about the features of this product below.
Seamless integration into your current process
Extract! CV parsing seamlessly integrates into the back-end of any CRM or ATS and can be added to the front-end of your career site through the ‘Apply with’ Widget. It can be used as a web-based application (in the cloud) or locally installed (on-site).
Data extracted from the CVs or social profiles can be customised according to your database fields or the requirements of your current recruitment or HR system. You can then build an optimum CV database ideal for sourcing, analysis and reporting.
For all existing integrations with suppliers of recruitment software, see our partner page. Would you like to know more about possible integration or documented technical specifications? Please contact us.
‘Apply with’ widget
Extract! CV parsing comes with a widget that you can easily add to your website. This ‘Apply with’ widget makes it possible for applicants to apply from any device (desktop or mobile) with their profile of choice. They can use their social profile from LinkedIn, Xing, Viadeo, Facebook, DoYouBuzz or Google+ or their CV stored on the device itself or on cloud services like DropBox and GoogleDrive. All information will be automatically structured and categorised according to your database’s structure. Read more about the ’Apply with’ widget.
Multiple field extraction
Extract! CV parsing is able to analyse each CV it processes in great detail. This enables the software to accurately identify and extract all the information contained in a document.
These fields include:
- Personal data, including name, contact information, social media links and profile picture
- Education, including qualifications, training and courses
- Work experience, including projects
- Skills, including languages and IT skills
- Soft skills
Textkernel’s extensive knowledge of the HR field coupled with its specialisation in cutting-edge machine learning technology makes Extract! the most accurate parser on the market. Textkernel’s quality team routinely monitors the accuracy of the parsers and its research engineers work constantly to further improve the performance of the various language models. Textkernel regularly releases updated versions of Extract!, and as a customer, you automatically benefit from these upgrades.
Multilingual CV parsing
Extract! CV Parsing is able to parse CV’s in a wide range of languages. It automatically identifies the language of the CV you want to parse or the social media profile you have acquired. Extract! is available in a constantly growing range of languages. We currently offer support for the following*:
* Is the language you need missing from this list? Contact us to discuss the current status of your required language.
Normalisation and ontologies
Extract! CV parsing is able to map extracted values to (hierarchical) taxonomies. One of our key strengths is to quickly build custom taxonomies for specific clients or industries together with our partners and customers. We also offer standard taxonomies. The software can be configured to do automatic taxonomy updates, so our parsing software and your back-end are always in sync. Besides mapping extracted values to taxonomies and standard skills extraction, Extract! is able to do fuzzy lookups in the text of CVs based on predetermined domain- or industry-specific lists of skills.
File types and data output
Extract! can handle a wide variety of commonly used document formats such as DOC, DOCX, PDF, RTF, HTML, TIFF, TXT, XML and EML. It can also process image files such as hard-copy CV scans. In addition, the software is able to extract data from social media profiles, including LinkedIn, Xing and Viadeo. Extracted CV or social media profile data can be modelled in any format you wish. Textkernel supports all current standard formats, including HR-XML (HR-XML Candidate & HR-XML Resume), out of the box.
Powered by machine learning
Machine-learning technology is at the core of our product development. In order to identify information from CVs and profiles, our extraction engine learns to create “rules” within its machine-learning algorithms by analysing large amounts of data. This technology results in parsing models that are more accurate and robust than parsing based on rules devised by human beings. Furthermore, new languages can be added more easily. For further information on our machine-learning technologies and our recent and exciting usage of Deep Learning, see the technology page.
For Textkernel, privacy is of the utmost importance. It is possible to use our software both as a web-based application (in the cloud) or locally installed (on-site). For the web-based version (SaaS—Software as a Service), where data transits through our servers, we comply with a strict privacy agreement. With each new customer, we sign a data processing contract ensuring that data is not distributed to third parties or used in any other way than agreed upon. Read more about data privacy.