HackStack

Project Info

Coffee Compass

Devpost

This project did not submit a demo video on Devpost.

Inspiration

Searching for your next coffee haunt can be time-consuming. Browsing Yelp or Google Maps involves a lot of filtering out Starbucks and McDonald's. Lists of the best coffee spots around town are scattered across various blogs and articles. And determining whether there are enough outlets or public WiFi available is simply a gamble you have to take. We wanted to make it easy to find ALL the information about your next coffee shop - including the little things.

What it does

Coffee Compass is a web application for coffee newcomers and enthusiasts alike, providing a large variety of recommendations for places to visit and can handle specialized questions so that you can find out anything about coffee and how you want to enjoy it.

How we built it

We scraped a ton of information from review sites, blogs, and informational sites and compiled it all into a corpus for reference. We adapted an LLM to be able to answer questions related to coffee shops and their specialities with a sense of personality as well as dynamically recommend places according to user info. The web application was designed with Figma, incorporating a chatbot conversational experience and passport incentive system to engage with users and keep them returning to Coffee Compass.

Challenges we ran into

Verifying information across different sources was super important. LLMs are more likely to hallucinate if they don't have accurate, verified information. It's a huge bummer to go to a great coffee shop by the beach, only to realize it doesn't have WiFi for your nifty MacBook Air. For the remote working crowd, even little things like outlets can matter. Being able to answer these questions gives a lot of value.

Accomplishments we're proud of

Prioritizing design and making sure the user experience is something that a user would really enjoy using. We really liked gamifying some aspects to reward coffee aficionados for their passion.

What we learned

Figuring out user workflow can be challenging, and there are lots of ways to tackle problems. Different kinds of users have their own problems, and that can potentially lead to design conflicts.

What's next

Build out the gamified passport experience so that visitors who visit coffee shops feel rewarded for their passion. Continuing to build out an extensive database of features for coffee shops that people care about and getting enhanced visuals for every coffee shop, so that users can really feel like they can explore the coffee shop before heading out there for a cup of joe.

Analysis

Compare with all teams

View

Metric

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only
  • CSSIn code
  • FlaskIn code
  • HTMLIn code
  • JavaScriptIn code
  • OpenAIIn code
  • PythonIn code

6 of 6 appear in the indexed code.

AI coding agents

No AI coding agent signals were found in this repository.

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

11 KB

Source files

6

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.

0 stars