sign-of-fourier/cfpb-complaint-enforcement

Used Perplexity computer to create agentic pipeline to predict CFPB enforcement actions

1 best points · 0 comments · 1 Hacker News post · 2 GitHub stars · View on GitHub

Social preview of sign-of-fourier/cfpb-complaint-enforcement

From the README

CFPB Enforcement Action Predictor An AI agent (Perplexity) autonomously built a Bayesian Optimization pipeline that predicts which companies the CFPB will take enforcement action against — using nothing but public consumer complaint data. The agent was told nothing about BoTorch. It independently chose the library, selected MixedSingleTaskGP as the surrogate model, and used LogExpectedImprovement as the acquisition function. Total cost: $200/month Perplexity Max subscription. No cloud compute, no GPU cluster, no research team.

Hacker News discussions

Show HN: AI agent autonomously used BoTorch to predict CFPB enforcement actions (1 points, 0 comments, by mshipman, 2026-03-16)