ThinkBoard

Inspiration

Brainstorming encourages us to utilize collaboration to solve problems and generate innovative solutions. However, it can also take a lot of time and resources to fully utilize brainstorming. In fact, that is the very problem our team faced as we were coming up with our project idea. We had spent many hours brainstorming, filling the whiteboards with ideas, but we still struggled with deciding the right direction to take. But, then it hit us that we could turn this problem into our solution! So, we decided to make a brainstorming app called ThinkBoard. The goal is to provide a platform that makes it easy to use smart technologies to improve and maximize brainstorming, helpings others to see the potential in their ideas and bring them to life!

What it does

ThinkBoard is an online web application that leverages the visual layout of mind-mapping, organizes your ideas, and leverages Natural Language Processing (NLP) using Cohere's models to provide predictive suggestions to plan your ideas based on your descriptions you type in. Furthermore, our platforms uses this AI technology to views summaries of your ideas, classify them into categories, and provide relevant suggestion's to narrow your ideas, allowing for further researching by clicking on related topics until the most closely relevant ones. This is all displayed in a concise and organized manner to manage your different brainstorms.

How we built it

All of our team members contributed their diverse skillsets to rapidly build our product in the following ways:

• Frontend: Developed a dynamic front-end application utilizing HTML, CSS, Javascript and the React Framework

• Backend: Incorporated Flask and Python to manage and create API endpoints

• Databases: Use of CockroachDB Serverless databases using SQL to manage data manipulation and store various types of data

• Machine Learning: We incorporated Cohere's NLP models and machine learning for processing data and real-time text analysis. Also used Open AI's GPT-3 for text generation

Our team worked in parallel between frontend, backend, data science and design, allowing us to build a strong working product that stores and loads information effectively, collecting and parsing data from multiple end-points. We also strived to create interactive website with animations and gestures.

Accomplishments that we're proud of

We were very pleased to see the process of seeing our idea come to life. At first we were worried about if we would find the right idea after so much time. But, we learned how to make use of the situation to try something new, and as a result we built something we are very proud of! We were also very excited when we saw how our program and it's accuracy given the models we used. It was also great to see our team work as we harnessed our different skills to produce our Full-stack application.

What we learned

We have learned that even the best ideas can spark from unexpected places. We didn't know where our ideas would lead us but with the right guidance we were able to find a project were we could all leverage our strengths. It was a great reminder to be willing to try new things, be passionate, and enjoy the process of learning!

What's next for ThinkBoard

Another feature that we would originally wanted to include is Speech-to-Text capabilities. This would allow our application to analyzing conversations in real time, which would play a huge rule in group collaboration tools. Other features include building a login-service to be able to save your plans. Additionally, with our backend system, we could it to save records or email them. We would also like to spend more time training the data to provide optimal results for searching for ideas in our app, as well as integrate other results such online resources or reference images to help with the brainstorming and prototyping process.

Built With

• api
• backend
• css
• data
• design
• figma
• flask
• frontend
• html
• javascript
• natural-language-processing
• open-ai
• python
• react
• sql