One of our teammates had spend his formative years in French immersion, while another through IB and into first year of undergraduate studies. Though, due to years passing by without sufficient practice and reinforcement, the had forgotten or found it hard to recall certain vocabulary. We wanted to remedy this by ensuring that foreign language speakers don't lose familiarity with important vocabulary that will be beneficial for strengthening their communication skills within those respective languages.
What it does
Autolingo opens up whenever a user opens a new tab on Chrome. Our app displays a word from a foreign language along with its equivalent word in English. Our idea is that through constant interaction with a foreign language's vocabulary, you'll retain your knowledge of the language for longer. In addition to consistent exposure to the language, our app occasionally quizzes the user about a word, so that the user can determine if they are properly maintaining their language skills.
How we built it
We built this app as a Chrome Extension. This involved using React JS for front-end, and using MongoDB Stitch for our back-end. Our database kept track of words that supplied our front-end tab. Additionally, Yandex's API was used to translate our practice words back to English.
Challenges we ran into
- Sleep. This was a project where we worked in a fast sprint-like setting to maximize the functionality of our product within a limited time constraint of the already limited 36 hours.
- Although few of our team members were acquainted with MongoDB/mongoose/graphQL, we were unfamiliar with MongoDB Stitch, so we had some trouble picking it up initially. However we were able to translate existing knowledge to make things work ultimately.
- After exploring many different ideas and pivoting between them due to constraints or blocks, we eventually started on this idea late Saturday night and into Sunday morning. The challenge/regret would be provided we had started earlier and had the opportunity to utilize more of the 36 hours provided, perhaps we would have been able to focus on more core functionality and even more distinguishing value propositions.
Accomplishments that we're proud of
We generally completed the MVP we set out for, even after pivoting from a completely different initial idea.
What we learned
- A lot of us, who are generally back-end developers, learned a lot more about front-end development, and gained an appreciation for it along with learning about MongoDB Stitch in its database capabilities but also having a chance to briefly explore the server-less stack.
What's next for Autolingo
- Additional Languages
- Questions that are tailored towards the user's weaknesses (i.e. ask more questions from areas of the language where they make more mistakes)
- Gamification, e.g. points, streaks, group leaderboards, experience points.
- Machine Learning with either pretrained/self-trained datasets (if there are no existing ones that fit well) to anticipate useful concepts to reach based on previous user experience/achievement.