To optimize the purchase of our next SUV, I developed an AI-powered data collection and decision-making pipeline. The system operated in several stages. First, I built a scraper to collect all car dealerships within a 100-mile radius from a major car brand's website. Next, I created a second agent to remove duplicates and gather Yelp and Google reviews for each dealership. This data was then cleaned and consolidated. I developed a custom scoring algorithm that combined Yelp and Google ratings, weighted by the number of total reviews and the distribution of positive and negative ratings. The algorithm ensured that dealerships with only a small number of reviews—whether extremely positive or negative—did not disproportionately impact the final ranking. Each score was normalized to create a single composite score for each dealership. I then searched dealership inventory for our exact vehicle configuration—color, trim, and features—and cross-referenced this with the composite scores and distance to produce a ranked list of best options. Finally, I retrieved the sales phone number for each dealership and created a call script to efficiently negotiate the best out-the-door price. Using this approach, we purchased our SUV for significantly less than the brand's own MSRP. The entire workflow is modular and repeatable, making it easy to reuse for future vehicle purchases.