AI Solutions

During a CV upload, a candidate mistakenly saved and uploaded the CV editing web page (in .mhtml format) instead of downloading and uploading the final document as a PDF. Currently, this type of file can be uploaded without any validation, even though it cannot be parsed by HrFlow. Could you explore a solution to either: notify us (the platform) when such a file is uploaded so we can proactively reach out to the user, or implement a way for this file type to be parsed if possible. Thanks!

We would like to gain more autonomy in handling situations where there is a discrepancy ("delta") between expected and retrieved profiles in the Parsing, Searching and Scoring modules. Currently, when profiles are missing from the results, we do not have direct visibility or access to identify which ones are excluded or why. We have to ask to your team to provide some answers. We are requesting the ability to directly access or view missing profiles —those that should be included based on input criteria but are not returned — so we can independently investigate, gain efficiency and take appropriate actions.

Currently, when a consultant uploads a CV in a single-page format with an unusually large height, the parsing process fails (likely due to confusion caused by merging what would normally be multiple pages into one). Since this format is relatively common, could you train the parser to handle it correctly and fix the issue? Thank you in advance.

Sometime the candidate make mistkaes and the mailto link is linked to text that is not an email or completly different from the text written in the CV. Could you check that when the parsing is running and take the text written and not the linked link? For example a CV pdf with a wrong email found in the metadata . Thanks

We recently came across an issue in the CV analysis API where there is a limit of 10 pages per CV sent. This ends up affecting the operation as in many cases we have CVs with more than 10 pages. I would like to request that you increase the limiter to 20 or more pages.

Currently, the field "description" present in an experience and an education currently contains data that is already contained in other fields, like the title, the dates, the company... Remove these duplicated information from the description field, in order not the have a same information present in two fields at the same time.

Take the level of langugage and not only the title or the name of languages

At the moment languages that are listed in CVs are extracted but the respective level is not (always resulting in undefined). It would be great if typical ways to write language skill levels could be parsed as well. I attached four typical ways levels are displayed in CVs.

Keep the structure of sentences and their indentation levels from the original CV + merge multiple lines if they are in the same sentence.

Structure a whole not structured job offer within different fields (company history, requirements, advantages, etc.) automatically.