Team 36 is honoured to submit the devpost regarding “MapleConnect”, an Android/IOS application designed to give assistance to the under-represented community that is the immigrants to come to Canada.

One problem that the team identified was that because of the COVID-19 pandemic, many immigrants were blocked in their country of origin, unable to come to Canada because of the closed borders, even though they have been accepted by Immigration Canada. This gave us the idea to try and give these immigrants a taste of Canadian experience beforehand. This will be done through access to news of your province of destination, many resources to accommodate to the Canadian life, but most importantly, Canadian mentors.

The main feature of the application, the mentorship program, will be based on a K-Modes Unsupervised Learning model revisited for the needs of our idea. The model will give to each mentor a given number of mentees that are most likely to share the same hobbies as them, in an effort to help them build a meaningful relationship. The mentee would gain insights on what life in Canada is all about, some tips and advice for the winter for example, gain valuable friendships, and possibly useful job connections! On the other hand, while the mentor would gain from the same social aspect, he or she would also have the opportunity to discover or learn more about a different culture, a different country.

The principal risk with this idea is that there may be more mentees joining the app than mentors, but we tried to mitigate that risk in different ways, not only by more promotion efforts towards mentors. In fact, the way our algorithm will work is that each monday, the clustering will take place and mentees will be assigned to a mentor, but only for a week! Thus, already existing mentors can be assigned to different mentees through time. We will also make sure to let mentors know that they are not expected to kill their free time answering the many questions their enthusiast mentees will certainly have! Moreover, each mentee will only have access to 3 mentors through his experience on the application, 3 weeks that is. What if they love the concept and want to continue? They sure can! For a fee of 0.99$ a week, they’ll be able to continue getting paired with new mentors. This fee option will allow the organization to generate extra cash flow while controlling the number of mentees for each mentor. For those who don’t want to pay the price, they will always have access to group chats that they can join, based again on their hobbies, origins, etc...

The organization is set to launch on January 1st of 2021, with its projected financials described by the table below.

While it is certainly very difficult to produce a financial projection during times of high uncertainty such as these where we’re living, it is even more for non-profit companies that are highly dependent on fundraising and other similar sources of financing. To produce a financial forecast that would describe best how the business would work, we decided to take a pessimistic scenario, but even then, the results are quite impressive. Our first ever source of capital would be a donation of $1500, an investment we as a team are willing to put to see our project take form. We also expect to be able to raise $500 in sponsorship during the next year, a figure on the lower end of what we believe is possible, and that would have the potential to grow exponentially. Indeed, more and more big corporations are looking for efficient ways to improve their brand perception, to attract new customers, increase employee satisfaction, and recruit rare talents. It happens that sponsoring a non-profit organization is one of these ways, and probably the best one when it comes to the last point, the majority of immigrants being accepted under the Skilled Worker program, that requires higher education. On another note, using the estimated figures regarding our client base that we’ll discuss in the next section, and conservative assumptions regarding in-app purchases (assuming for example that less than 4% of mentees buy at least 1 more week) , we come up with the third type of financing.

Regarding the operating costs, we would work on the following basis: a fixed cost each year for a hosting subscription in Amazon Web Services, at the very competitive type of $0.023 for each GB and each month. The next priority would be given to Selling, General & Administrative costs which represent, in the first year, costs of incorporation and pattenting for example. The remaining cash will be used for promotional purposes. In 2022, even if we do not expect any SG&A expense, we still consider a 5% of revenue figure to be available in case of extraordinary costs. However, change happens as soon as 2023, when our financials would have improved enough for us to try a new challenge, hiring contractual “trained mentors” that would have gone through an educational program designed by ourselves to be able to advise users in the best possible way. This is represented in our financial model by a sudden spike in SG&A costs as a percentage of revenue, that continues to stay big as we keep developing this promising program.

Our motto is: “Canadian-based, Human-centered”. It perfectly sums up the intention and the direction of the venture. We have deep Canadian roots, but we are dealing with humans from all around the world, no matter where they are, or where they come from. Our target market is all immigrants accepted in Canada. However we can segment the population into the 4 different categories of immigrants including skilled workers, students, refugees and those who have been sponsored by those in Canada. We chose to specifically target skilled workers as they represent the greatest proportion of immigrants coming to canada along with students who are approved to arrive for PR or work-permit status. Our TAM (Total Addressable Market) is about 284,387 people according to Statscanada. Our SAM (Serviceable Available Market) is about 250,000 users according to competitive statistical sources. Our SOM (Serviceable Obtainable Market) is about 222,000 users according to Stats Canada. Skilled workers who are most likely to use our app as they tend to be more anxious about visiting Canada and are often a younger population which makes them more likely to engage in an app and use the social aspect of the app via the mentorship program. Additionally our primary target of skilled workers are at higher risk of feeling the sense of isolation, allowing them to be more likely to use MapleConnect. By being more active on the application, they will be invited to pay a small fee to keep using the service of the applications after the 3 weeks threshold, to insure that new immigrants get paired with the mentors, and to avoid overcrowding in the application as we stated before. Also, the companies and economic sectors that are hardly affected by the pandemic and willing to attract labor would be really interested in getting access to such a large pool of candidates. The application indeed features a career section where they can have access to career advice, and job postings. We can charge a premium for the companies willing to look for their jobs to be listed on the application. Although a competitor service does exist:; MapleConnect provides the same service as in terms of resources and guides; however, MapleConnect adds the concept of personalization through customizing feeds and content and emphasizing mental support through the mentor program. Further, our platform provides a way for immigrants to actually meet and interact with other people similar to them and form new friendships forming the building blocks for settling down rather than just information. The app-format allows for better accessibility which is unique to the market as there are no other apps providing aid to immigrants which makes the information much more organized and user-friendly.

Now our first thought once we secured our idea was about the implementation. We wanted something that was accessible, easy to use, and high performances. For the simplicity of usage and accessibility, we decided to develop a cross-platform mobile application available for both IOS and Android. The engineer of our team responsible for the front-end didn’t make it to the end of the competition, we had to change our plans, but we still managed to develop a user interface (UI) design for the application using Figma, a prototyping application. The prototype is nine pages long, and goes over the main functionalities of the application. Such features include a social media component, a resource list, and a news feed. It has a strong focus on Canadian life and values, in order to provide a better immersion experience for the immigrant before their arrival. The choice of the blue color is due to the psychology behind the color. The Blue color has calls to mind feelings of calmness or serenity. We hope that through this attention to details of the UI, the person behind the screen can feel a little bit less anxious and stressed, which unfortunately happens a lot for immigrants due to their current situation and uncertainty. For the second main point, the intelligence of our software, we decided to develop a Machine Learning algorithm that would fit this problem. The reason being that we wanted the algorithm to learn from its mistakes, and to get better with time. We looked into many algorithms, but our use of Categorical data made our task a lot harder. Very few mainstream algorithms met our needs so we read some scientific papers in order to look for a very specific solution. We wanted to classify the mentees of MapleConnect and assign them to a mentor based on many categorical variables such as age, province of destination, ethnicity, hobbies… After a lot of research, we finally decided to implement a K-modes classification algorithm in Python, which is what best fitted our needs. We were able to minimize the cost function based on Hamming distance, allowing us to get the best recommendations for each individual in the list. The clusters are also changing based on the number of mentors and mentees, which allowed us to have very dynamic pairing within the application. Even if we were not able to implement the algorithm we developed with the front-end as we decided to go for a prototype demo, the result provided by the algorithm exceeded our expectations, and the allocation of the people registered in the application was successful.

Finally, regarding the political environment, the use of MapleConnect may allow for the promotion of immigration to Canada due to the increased amount of support received to immigrants. This would substantially boost the Canadian economy through increasing our labour force while also providing more opportunities for new businesses and expansions. Immigrants represent a major component of the economy, it is estimated that starting in 2031, more than 80% of our demographic growth is projected to come from immigration, compared to about 67% in 2011 (Statistics Canada Catalogue, 2012).

Built With

Share this project: