I built this app for the meta worldwide hackathon with cerebral valley. I used muse code and muse family models for the first time to execute the project. I received access to a few hundred dollars of compute and inference resources to build a project. Thanks to the sponsors.

This is an open source (MIT) front end for buying a car. I have curated the vehicles to support a decision that will be easier to survive without financial regret.

Some of the strengths of the app are bypassing all of the bs associated with a traditional car buying experience, which has tons of cognitive overload and not enough pure salient information.

Low Key prioritizes - your effective monthly spend on fuel and the car loan - the positive equity in the car so that you can sell something you own for a benefit when you’re done - safe and durable vehicles with a good record of reliability’s - competitive best practices for rating deals between dealerships

It is OSS and I would love to see anyone’s fork or stars to contribute better ideas than my own.

Advanced features include vehicle ID by VIN, far flung low price finder, and time series foundation modeling of fuel prices by zip code and week.