My DSN Impact Story

by Franklyn Ogbonna

I started to get new feeds from Data Science Nigeria on my Linkedin timeline from the last quarter of 2019. By merely looking at the name, I got attracted to it and decided to follow up and see how I can benefit from their resources and events to help me transition into a Machine Learning Engineer. I have always had the burning obsession to launch fully into Artificial Intelligence/Machine Learning, but was still fighting with the inertia to kickstart it.

On 16th of July 2020, a linkedin user, Jibowu Omotayo reached out to me, introducing to me the AI Health and wellness Hackathon organised by DSN. I immediately picked interest and we formed a team. Prior to this time, I have been fiddling with some Machine Learning (ML) models from various online resources of which never left the confinements of my local jupyter notebook.

We finished work on our project, built a classification model to classify any food image uploaded to it and tell its nutrient class, constituents and as well suggest supplementary meals according to one’s age, gender and health conditions. Everything seemed perfect on my MacBook Pro.

Then came the headache. How do I serve our application so that it would be usable and people can appraise our handiwork? My first idea was to put it up on heroku. I tried with the regular heroku deployment with no luck owing to slug size limitation. Our model’s size plus other accompanying resources made the app’s size considerably large to host on heroku free tier, and we were not ready just yet to spend a penny on it. It’s just a hackathon project, we thought. We did not know our fate yet. I spent a good 48 hours surfing the web for solutions and trying various implementations without success. I became sad, not knowing how to explain to my team and the world that I could not host our project that we have successfully built and works well on my local PC. Rest assured that, to every problem, there’s a solution, I continued on the 3rd day and eventually had to containerize the deployment with Docker and voila! It was a Eureka! I successfully deployed a working model for the first to a production environment, ready for public use, courtesy of the hackathon organized by DSN. I was extremely happy, not expecting anything than just being happy for the great feat.

A month passed and it was time to announce the winners of the hackathon. You guessed right, my team came top first in the DSN AI Health and wellness Hackathon, 2020 and won a cash prize of $4,000. I could not believe it. It was surreal to me and my enthusiasm and drive to pursue my ML career once again soared.

I’m really happy and grateful to DSN for the opportunity and events I have been part of and hope to stay more active in the DSN community and also give back to the society.