Recall

Inspiration

Often we have countless tabs open while studying or programming can't seem to find that one tab we had open

What it does

Using a word2vec machine learning model, we search through the text of your past sites to match keywords, being able to find information without remembering text word for word.

How we built it

We used a python flask server to determine how close sites are to search queries. The front-end was handled in javascript and bootstrap html.

Challenges we ran into

It was difficult to connect everything together and run a stable server on GCP

Accomplishments that we're proud of

Using Microsoft Azure Cognitive Services, we were able to search for a video visually and return a timestamp

What we learned

There is a strong trade-off between the accuracy of searches and the time

What's next for Found

Optimizations for the natural language processing and a more stable server.

Built With

• bootstrap
• flask
• javascript
• machine-learning
• python