I seek Project Partners

Our AI Robot evens out the male dominance in recruitment to the UN and to International Organisations

Digitalization and Technology are both under-utilized when it comes to solving the diversity and the gender equality challenges we face.  Now, Impactpool applies for public funding to take their AI-algorithm to its next level and to establish the Fair recruit and search console. A recruitment algorithm trained on impact jobs and on global impact talent profiles to better handle bias related to culture, age, name, nationality, gender, sexual orientation, disabilities, and more.  Now you can join force and build the Fair recruit and search console, exclusively built for the impact sector. Just submit the form at the end of the article and let us together hit the ground running. 

The talent funnel to the international organizations is men heavy 

It is no secret, the average applicant sourcing funnel to the United Nations consists of up to 3 times more men. In a random data sample of staff in these organizations, the white men are over-represented, he is the norm. 

Women are under-represented especially at the senior level, but the truth is that women are generally under-represented at all seniority levels, and it becomes more obvious when studying women from, Middle East, from Asia, and from countries South of Sahara. 

It is also generally rare to find staff members from countries affected by recent wars, crises, or from some specific nationalities and minorities. 

Other broader categories that are also rare to see are young professionals (30 years old or younger). In fact, if one does not include positions funded by donors (Junior Professional Officers, JPOs, and similar), for the whole UN system out of 38710 international professional staff, only 7 are under the age of 25 years (CEB dataPersonnel statistics 31 December 2018). It is also rare you see staff members with a physical or mental handicap or staff openly non-heterosexual.

In our project scope, we define all these categories as under-represented groups, in a local context these groups would probably fall under the category of protected groups. As International Organizations and the United Nations serve as role-models for local societies, an unequal or non-representative workforce is likely to hold back positive change related to diversity and gender equality.

A report followed by a successful BETA release gave us the confidence to start our journey to break established norms

In 2016, only one year after Impactpool was founded, we released a study on the price women pay for their UN careers - Are Women Paying a Higher Price for a UN career. The study presented data in a way no one else done before and the Secretary-General of the United Nations reference to the report in the UN System-Wide Gender Parity Strategy

During 2018 we continued the ambition to make international organization equal and invested significant funds to release a BETA of the first-ever AI Impact title mapping robot. An AI robot that understands the job titles of the Impact sector.  Over the past 12 months, our AI robot has been trained on approximately 300’000 impact job titles, and the result is stunning. Check out our robot here. What the AI robot already does is to understand English job titles and translate them into matching expertise areas. This is a way to support individual applicants who are not used to present themselves using the type of CV templates that western societies are using and template that has become norm also for international organizations.

How can this be used to break established norms?

Our goal is to use AI technology and train a recruitment algorithm on global talent data to establish a  Fair recruit and search console tailored to the Impact sector. The AI robot will acknowledge different ways of presenting oneself, hence treat individuals equally and thereby support organizations to avoid bias related to culture, age, nationality, gender, sexual orientation, disabilities, etc.  There are some known examples of when AI recruitment projects failed or even worsen the situation for under-represented (protected groups) groups, but our project has another goal and has therefore much more positive chances to succeed. Most of the past AI projects have been focused on finding one best talent, but with the type of historical and current applicant data available, that ambition is not yet realistic.

Our AI robot will not be assessing candidates' suitability. Instead, it will use an innovative ranking algorithm to check for highlights in their applications -  we have named these highlights markers. A marker can entail different things, it could be candidates having experience from a certain country, skills in a certain language, specific certification or university degree, etc. The AI robot will be trained to seek and find markers in an innovative sophisticated manner with the purpose to break the Norms of the current CV, and to allow different cultural ways of presenting a career and professional background.

The ranking of candidates will purely be based on these markers.  In Phase 1 we will develop the algorithm and train the AI robot to understand:

  • the classification structure

  • grades and contract types

  • job locations (family/hardship)

  • languages

  • education level, education profile, and institution

  • if a job is local or international

Today, larger organizations are having several hundred thousands of application records, but current recruitment systems only allowing limited search capabilities. To speed up the fulfillment of goals such as gender parity, organizations will through this project be able to use our AI robot to quickly complement their current talent funnel by using our AI robot to search historical application records and reach out to talents one has missed out on in the past.  

Why is a Swedish Startup best equipped to move this project forward?

Sweden as a society is always top-ranked in international rankings on gender equality and innovation. In this project we allow an agile startup to mix these two skills.

Impactpool was founded in 2015 by HR professionals having a long career in international organizations. One of the founders, who will also be the project lead for this project, led the implementation of UNDP's recruitment system, the second largest recruitment system implementation in the history of the UN. Already Impactpool works with more than 700 impact organizations and has built a pool of close to 450'000 impact talent profiles from 200 different countries. Impactpool's network and global pool of impact talent profiles are unique and something very few (or in fact no one) can compare with. 

Is this project idea something that gets you going? 

We are now seeking International organizations that want to join forces in this exciting journey and partner with us in the core project team.  In Phase 1 partner organizations contribute to the algorithm design. In Phase 2 partners support data mapping and in Phase 3 partners will be the test pilot. The project goals are to:

  • Offer partner organizations to use the AI robot to search for talents from under-represented groups

  • Allow partner organizations to build an API so that the AI robot can be used for fair screening of new talents and to search for talents from under-represented groups in historical data.

Do you want to be part of our AI project?

Are you interested in having your organization joining this project in the core team? Please fill out the contact form below.