The first time I implemented Vamana from the DiskANN paper, my approximate nearest neighbor index was slower than brute force. On tiny test fixtures, brute force took 0.27 ms per query. My Vamana implementation took 22.98 ms. That sounds absurd. ANN exists to skip work. The problem was not the algorithm. It was how I mapped the paper's abstractions to actual data structures. The DiskANN pseudocode
For years, I called myself a web designer. Then a developer. Then a digital consultant. None of those titles ever felt quite right. Because clients weren't just asking me to build things. They were asking me to solve problems. Slow sites, broken checkouts, confusing navigation, teams that couldn't figure out how to update their own content. That's when I realized what a technology solutions profes