Rafiqi

Rafiqi is a platform that leverages AI for facilitating the lifelong employment of resettled refugees

Today's world count more than 25 Million refugees resettled in host countries. In 2015, more than 1 million refugees sought asylum in Europe alone. When fleeing conflicts, refugees have one goal in mind: thriving in a new place they can call home. While their first efforts are geared towards settling down and legalizing their status, their attention is then directed towards lifelong employment. Although many governmental, non-governmental and private entities are working with refugees towards enabling their employment, our exploration of the refugee ecosystem for more than 1 year allowed us to identify the following challenges:

• lack of a single platform where newcomers can navigate the range of opportunities available to them, and where NGOs, universities, employers and volunteering mentors can access and filter refugee talent
• Newcomers’ unawareness of opportunities and of the right opportunities results in them being unemployed or being stuck in low-wage low-qualifications and often short-term jobs. According to Nuffield foundation, 63% of refugee men working in the UK are overqualified for what they are doing.
• While there are significant efforts by governments and NGOs to match refugees to opportunities, scaling these efforts cannot be done manually given the number and diversity of both refugee profiles and available opportunities. opportunities include jobs, job coaching programs, technical trainings, trainings in soft skills, mentorship, and accredited education.

Rafiqi was introduced with the following goals in mind:

• Providing a single platform for democratizing refugees' access to opportunities, and for NGOs, universities, companies, organizations and volunteers to easily access and filter refugee talent

• Leveraging machine learning for connecting refugees in real-time and in a customized way to the opportunities that are the most inline with their current potential and that will eventually lead them to high quality lifelong employment. The machine learning technique that we use is called decision trees. Our matching algorithm mainly consists of using different elements of refugee data, such as country of residence, level of education, education and work background, job readiness, career goals, digital and language literacy levels to better refine the selection of opportunities till eventually reaching the most suitable ones. 

Rafiqi leverages machine learning in a sphere where it is the least utilized, that is refugee aid. However, if we look at how diverse the profiles, needs and aspirations of refugees are, and how wide the spectrum of opportunities is, data intelligence and (semi)-automated decision making can bring a lot of benefits. 

While existing initiatives focus on providing a well-defined service to refugees (example: teaching them a language or how to code), Rafiqi allows them to navigate the range of services available to them and to be matched in real-time to the most suitable ones.

Rafiqi uses technology in 3 main ways:

1. using web scraping for collecting data about opportunities available to refugees

2. Automated clustering of collected opportunities based on info such as mode of delivery (online/onsite/both), location, theme, type (job/training/accredited education/mentorship...)

3. matching between refugees and opportunities is enabled through a machine learning technique called decision trees, where different elements of refugee data act as tree nodes and opportunities act as tree leaves. Feedbacks from refugees and opportunities providers about the matching output will eventually feed into the matching algorithm and make it better.

We currently have a MVP of the matching tool that is being tested by a dozens of selected refugees in London and Berlin. Over the next 12 months, we would like to do the followings:

1. After initial testing, pilot the revised MVP of the matching tool in at least 2 European cities with a strong refugee presence and among a selected group of urban refugees in Jordan

2. Incorporate wider audience feedback (users can provide feedback directly through the tool) in the UI and algorithm design of the web tool 

3. start prototyping a mobile app and a chatbot

Our first vision is to make Rafiqi the platform of choice for the millions of resettled skilled refugees fighting their way towards lifelong employment. To get there, we need to combine building intelligent scalable and user-friendly matching platforms with the ability to scale our reach and appeal to both our potential partners (NGOs/employers/universities/mentors) and our target group (refugees aged 18-35 with intermediate digital literacy and English levels). 

Our second vision is to scale Rafiqi beyond the refugee community to include local people with immigrant background living in difficult suburbs, as well as people working in soon-to-be automated jobs.

To attract and retain refugees, we plan to identify ambassadors among them who will be the first to use the platform and help us spread the word about it. Once we reach a critical number of users per major city and we incorporate enough feedback, we will use social media groups to scale our targeting and reach. Retaining the target group happens through incorporating their feedback, scaling and diversifying further the opportunities database, and featuring success stories.

we attract and retain partner organizations by providing them with a high quality low-cost and very quick access to refugee talent.

The typical refugee that can benefit from our platform is someone aged 18 to 35, with intermediate levels of english and digital literacy.

So far, during the idea phase, we have manually matched 30 refugees falling in this category and based in the Netherlands to opportunities, mainly focused on developing and advancing their tech skills. 

We used the learnings from the idea phase to develop a prototype of the matching tool which has so far been tested/used by 20 refugees.

we expect to reach around 1000 people in the coming 12 months, and hundreds of thousands within 3 years. As previously mentioned, we plan to reach our target group through word of mouth delivered by local ambassadors prior to using targeted advertisement through social media groups (refugees use heavily facebook groups to interact). We are particularly optimist about the word of mouth technique as we have already gathered a lot of enthusiasm and positive feedback from potential ambassadors who tested the matching tool.

Ghida has a PhD in Computer Science and ~7 years of tech work experience including working for Facebook as a data scientist. An immigrant from the middle east based in Europe since 10 years, she relates to the challenges of Syrian refugees and can address them in their native language.

Suzanne is based in Jordan and worked for 8+ years in PR and Marketing including working for Yahoo in the US. The fact that she is based in Jordan, a country with 2 Million refugees and hundreds of refugee-focused initiatives, makes it possible for Rafiqi to scale beyond Europe.

In the initial stages, we count on external funding to build our platform beyond the MVP and scale our marketing efforts.

However, a business model can be envisioned. Here are some ideas:

1. Sponsored Posts - Recruiters, training centers, and service providers can pay to advertise on the platform to reach more users (via custom targeting). For example, a training center can advertise coding or application-building courses to refugees looking for jobs at technology start-ups. 

2. Rafiqi as an outsourced AI service - Further down the road, Rafiqi's AI technology can be white-labelled by other job/opportunity matching platforms that do not have the resources or infrastructure to develop this technology in-house.

3. Charging a fee to employers for each successful job placement through Rafiqi

Supported by MIT, SOLVE can help us access renowned AI practitioners who can help us further develop the underlying technology behind our product.

SOLVE international recognition can give us more visibility and facilitate our connection to key partners that can help us reach more refugees and create more opportunities for them to make them ready for the workplace of the future (tech trainings, soft trainings..)

The 10k money grant awarded to the winner can also help us develop our product beyond the MVP so it can be better ready for the pilot and growth phases. 

 Main barriers we are facing include increasing our visibility, accessing monetary capital and making our product more intelligent and scalable.

Solve can help us through connecting us to worldwide key partners and bringing awareness to our project/product.

Solve can also facilitate our access to advisors and tech experts from MIT who can provide us with tech mentorship to make our product better.