With rent in Toronto breaking records, it is more important than ever for young adults seeking financial independence to correctly identify the right place and time to move out from home. Moving out from your parents house is one of the biggest steps students and young adults make to gain financial independence. However, rent prices of over $2000/month is commonplace in the city of Toronto making it extremely difficult for young people to live on their own. Hence we devised an innovative solution for this problem, which at its core is utilizing machine learning to asses the sustainability of students lifestyles, and offering them insights into the best area and time to move out.

What it does

Leveraging AI technology and the TD Da Vinci API, MoveOut provides personalized insights and aims to equip students with the information needed to make smarter financial decisions. Using machine learning algorithms such as distributional mapping and k-nearest neighbours, MoveOut analyzes data from over 10,000 users in 5 regions of the GTA to produce intelligent estimates and ensure the user makes the correct decision.

MoveOut analyses a users current spending across multiple categories across time and compares it to thousands of other customers to predict how their spending habits would be different if they were to move out and even move municipalities. For example, a student living at home in East York who is wanting to move out of home into North York will have different cost of living such as rent prices. Our AI projects an individuals spending habits to see how living in North York, out of home, would affect their costs.

Further, MoveOut also uses clustering and customer neighbouring models to match similar customers in the TD Da Vinci API who are successfully living out of home and in the municipality they are interested in. This comparison allows users to gauge how capable they would be in living out of home too.

How we built it

The implementation of MoveOut starts with a rough architecture comprised of 3 segments, UI, Backend, and AI. These modules were built with scalability in mind, and work independently, allowing for an efficient and customizable product.

The product leverages extensive API and framework usage for different and diverse purposes including the TD Da Vinci API, PyTorch, scikit-learn and pandas. The front end was built with React.js and spoke to a GCP-hosted Flask web service that ran both our servers as well as our Machine learning.

The data set used to build the ML models was retrieved from the TD Da Vinci API. The data set was compiled from the API by a distributed processes queue, allowing for a larger throughput and ETL process for customer data. Dimensionality reduction, feature extraction and data imputation were implemented to pre-process the raw data, fit for model use. This ETL was modularise to be ported directly into the backend, allowing the preprocessing of serving data to be consistence with how the models were trained.

The Machine Learning usage allows the product to return intelligent insight, and the use of clustering algorithms gave our dataset the feature to extract the cost of living in a particular area.

Challenges we ran into

One of the first challenges that we ran into as a team was learning how to properly integrate the TD Da Vinci API and how to properly process the data. With each API call returning 1000 users, and the dataset containing over 1,000,000 entries, the scale of the project grew quickly! We were finally able to work through the problem by leveraging cloud computing and cloud services (GCP) to process the data efficiently.

Furthermore, another challenge we ran into was the implementation/construction of our machine learning based REST API, as there were many different parts/models that we had to "glue" together, whether it was through http POST and GET requests, or some other form of communication protocol.

We faced many challenges throughout these two days, but we were able to push through thanks to the help of the mentors and lots of caffeine!

Accomplishments that we're proud of

The thing that we were most proud of was the fact that we reached all of our initial expectations, and beyond with regards to the product build. At the end of the two days we were left with a deployable product, that had gone through end to end testing and was ready for production. Given the limited time for development, we were very pleased with our performance and the resulting project we built. We were especially proud when we tested the service, and found that the results matched our intuition.

What we learned

Working on MoveOut has helped each one of us gain soft skills and technical skills. Some of us had no prior experience with technologies on our stack and working together helped to share the knowledge like the use of React.js and Machine Learning. The guidance provided through HackTheNorth gave us all insights into the finance industry, with many of Canada's largest financial institutions being onsite at the hackathon. Apart from technical skills, leveraging the skill of team work and communication was something we all benefited from, and something we will definitely take with us in the future.

What's next for MoveOut

Moving forward we see MoveOut expanding to other large cities like New York, Boston, and Vancouver as well as for students from smaller towns who may be transitioning after university. High housing prices are not an issue unique to Toronto, and we want MoveOut there for the students in need of aid. If one day MoveOut could be fully integrated with financial planning services provided by a bank, and could be provided in any Metropolitan area, our goal would be achieved.

Finally, we see MoveOut partnering with a number of large financial institutions that would like to use the service to better serve their customers, provide improved financial planning for students, and help them make better decisions. Partnering with a large financial institution would also give MoveOut the ability to train and vastly improve its model.

To summarize, we want MoveOut to grow and help students across the Globe!

Built With

Share this project: