Reddit posts related to Reddflag.
Companies spend hours combing through social media sites to learn more about users are experiencing their products. Tools exist for brands to engage with their customers on sites like Facebook and Twitter, but it's particularly difficult to interact with them on Reddit.
What it does
Reddflag ingests posts and comments from Reddit, looking for mentions of a company and words related to that company. If the content matches, Reddflag scans its text for sentiment (was the poster generally positive or negative) and shows it to the brand. Brand managers can then respond directly to customers through the app.
How we built it
Our backend is a django app hosted in a Google App Engine Flexible Environment. We use Cloud Functions to ingest and process messages from Reddit and Google's Cloud Natural Language API for sentiment analysis. Our frontend is built in React.
Challenges we ran into
- Time! Never enough time to do everything you want to do!
- Rate limits on Reddit (our test accounts don't have nearly enough Karma yet
Accomplishments that we're proud of
We're proud to have a working prototype of our project that works end-to-end with a single brand!
What we learned
We learned a lot of new technologies (particularly on the Google Cloud stack)!
What's next for Reddflag
- mark posts as "done" after they've been responded to and "reopened" if the original poster responds again
- integrate with Zendesk so companies can use this app within the context of tools they already know
- allow companies to sign up themselves!