Inspiration

We love great products. But more often than necessary, a product gets hyped before release and does not meet the expectations of the customers. We saw this with famous examples such as the Galaxy Note 7 (where the hardware failed spectacularly) or No Man's Sky, a game from indie developers that, after a very successful campaign, was a huge disappointment and missed many of the gameplay elements that were promised.

This is not fair, and we think we should give people an opportunity to secure their investments, if a product disappoints completely, while still being able to preorder if they want to.

What it does

ColdFeet lets you choose a product you have pre-ordered, and after we verify you have actually made the purchase, we secure your purchase! Lets say you have pre-ordered a smartphone on Siroop:

  • We calculate the premium based on historical sentiment data from Twitter
  • As soon as the product is released, we enter a waiting period where critical flaws can be detected by the community. This period may change depending on the product. We detect bad products using Sentiment Analysis and Natural Language Processing, using Social Feeds as Input. Currently we parse tweets that talk about the product.
  • If your product has a major flaw, and the general sentiment is very negative, we will re-imburse your purchase using a Voucher Code :)
  • If your product is good, we keep the premium, and our customers can be happy that their product fulfilled the expectations.

How we built it

To be flexible, agile and prototype rapidly, we chose NodeJS for the backend and Angular2 for the frontend. To enable interactive and real-time features, websocket technologies such as socket.io are used to transmit data. We store data in MongoDB and use IBMs Watson API to do the sentiment analysis and NLP.

Challenges we ran into

Correctly handling and streaming the continuous stream of tweets from Twitter was very challenging and led to some savvy workarounds. Matching products from a shopping API such as Siroop to tweets was not easy neither.

Accomplishments that we're proud of

We're especially proud of how easy it is to use. The application looks sleek and has a consistent design. The real-time feedback of the tweets and the sentiment analysis further improves the user experience.

Share this project:
×

Updates