Looking at doom and gloom everywhere, I guess 2023 is the productivity year! Imagine you can do 3x more with the same resources. Sounds intriguing but is it really possible? The answer is yes, because today you have a set of tools that were available only to major corporations a couple of years ago. Now, you can perform as 50 employee company while being only 3 of you.
So in this article, we will teach how to improve productivity in your hiring and candidate-sourcing operations.
What we will do:
Use a CV/resume PDF as a single source of information
Setup an automated data extraction from a CV PDF to your tool
Then we will categorize the candidate based on expertise level, seniority, and previous companies
Detect candidate profile (technical/business/financial/etc.) using skills from CVs
Configure the dashboard to see all incoming candidates in real-time
Setup openAI to get us a list of interview questions for the candidate
TL: DR; video
The tools that we will use are:
Dropbox or Google Drive
Optional tools:
LinkedIn
OpenAI ChatGPT
Slack
Our dashboard will look like this:
As now our information is presented in a neat view, we can also build dashboards.
Part 1: Collecting CVs PDFs to a single location
One of the key processes we need to establish from the beginning is forward all information to a single location. Everybody should have a simple instruction - if you receive a CV or export a LinkedIn profile to PDF, it should be saved to this location.
In our case, we will use a standard cloud folder as source.
Thus, these are the easiest ways to submit the information:
Save CV PDFs to cloud folder
Forward a CV to a special mailbox
Submit a CV PDF via form
Save PDFs to a cloud folder
It is very simple - when you receive a resume, you can immediately upload it to a folder called “Candidates Incoming”:
This step already creates a single source of information in your company or team.
Forward a CV PDF to a special email address
You can also create an email address like candidates@mycompany.com and then forward PDFs there. This solution would allow you also to get the information about a person who sent the CV PDFs.
After the file arrives to a mailbox, you can also save it to the same cloud folder.
Submit a CV PDF via Form
If you use tools like monday.com to manage your candidates, then you can set up a form for your team to load the PDFs:
Thus we are now ready with our single location for CV PDFs and we can proceed to automating data extraction.
Part 2: Extracting data from CV PDFs
In this part, we will set up automated data extraction from business cards. We will use Make.com as our main workflow management tool. You can replicate a similar scenario in other tools like Microsoft Power Automate or N8N.
We will use a trigger (New Email, New File or New Form Submission) to detect that something has been uploaded:
Now let's configure the Bitskout template. Go to Templates Library and search for a template called “CV analysis”. This template allows us to automatically extract information from any CV in any format:
Simply, click on “Use template” and then follow the steps to set up the plugin and you are ready to go.
Now let's add the Bitskout module to the Make.com scenario, choose Run Plugin for a File and select our plugin. To specify the URL you need to map the “Download URL” value from the previous Dropbox step.
Now let's load the file and test it. In Make you can run a module by right-clicking on it and choosing Run this module manually.
Thus, now anytime a new file is uploaded to a folder, the scenario will run automated data extraction.
Part 3: Categorizing Candidate Profile
The target of any process is specialization thus your current setup is optimized towards a certain ideal profile that fits into that categorization.
Depending on the position you've published you might have different criteria. For this example, we will create two categories: seniority level and experience level:
Seniority: Contributor, Mid-level Manager, Executive
Expertise: junior, mid-level, senior, staff
Now you can create your own categories, that's the beauty of the Bitskout A.I. - it works based on your examples.
Before we start adding examples, we need to decide the key information element that we will provide to A.I. as an example.
This is where Bitskout A.I. help you build that. We'll do that by adding examples. Let's go to Create Plugin -> Understand -> Categorize.
Let's add examples:
All right, press Finish and then press Next to set up the output. Once you select Make.com, you can go to the next step. If we use Make.com or Zapier, you don't need to set up a separate output.
Now our plugin is ready, let's add it to Make.com scenario via Run Plugin For Text module:
Once you've chosen the plugin, in the Input text you need to select the JOB_TITLE from the previous step. This way, once the previous Bitskout plugin extract the Job title from a CV PDF, it will be then mapped to the next step.
Part 4: Creating the Candidates Dashboard
The next step is to create a dashboard where we see all the incoming candidates. Using this dashboard you will be able to analyse the current situation. You can also add position ID to sort the candidate per position.
This step allows start creating your own candidate database.
First, we need to set up our board:
Now let's configure the scenario in Make.com to create the item. We will use monday.com Create Item module:
Once you add the module, you will be able to map columns in your monday.com board with the values extracted by Bitskout.
As a recommendation, we also advise adding a CV PDF file to the file column.
Now, once the file is loaded to the Dropbox folder, we will be able to analyze it and create an item in monday.com with all required information.
Using monday.com allows us to build dashboards:
This way you will be able to visualize information in a real-time manner. Also, using cloud tools allows you to search for candidates for other positions, create keywords and match them with job position descriptions, etc.
Bonus: Using openAI to send interview questions in real-time
As a bonus, let's add some generative A.I. to help us with candidate interview. As you have introduced various categories, you can tune the interview questions based on various criteria.
To do that we will use openAI ChatGPT module. openAI works with prompts, thus, what we are doing is asking to create a list of 5 interview questions for a certain position in a company. This way our list is tailored to a certain position. And then we send it as a Slack message.
This will generate the following output in Slack:
Conclusion
In this example, we wanted to show how having just a single source of information you can build a very sophisticated and efficient process. You need just four tools for that and 1-2 hours of your time.
And you will be able to absolutely transform your recruitment processes making them fast and productive for your team and your candidates.
And most importantly, you can do it by yourself.