Wait, this is sick. Which model did you guys end up using under the hood to do all the classification for this? How did you guys think about false positive rates?
Really like the design choice of keeping everything local (charts, research, indicator code). Privacy-first + open source + agent-native is a combination I haven't seen anyone else ship for markets. Following this…
Wait, this is sick. Which model did you guys end up using under the hood to do all the classification for this? How did you guys think about false positive rates?
Really like the design choice of keeping everything local (charts, research, indicator code). Privacy-first + open source + agent-native is a combination I haven't seen anyone else ship for markets. Following this…