Teams know churn is a problem but stall at building a first model from scratch. This automation ships a working baseline you can act on.
sonnet1 weekPythonscikit-learnPostgresSHAP
Claude
75ROI
76Scale
$2.7k91Saved
ROI for
README.md
Why this subagent
Teams know churn is a problem but stall at building a first model from scratch. This automation ships a working baseline you can act on.
It pulls account history, builds standard features, and trains a gradient-boosted model, then scores live accounts by churn risk. Each at-risk account comes with its top drivers from SHAP, so the success team sees not just who might leave but the reason behind the score.
How it runs
Used at step 01 to kick off the pipeline.
Write
Used at step 01 to kick off the pipeline.
WebFetch
Used at step 01 to kick off the pipeline.
WebSearch
Used at step 01 to kick off the pipeline.
Build features for tenure, usage trend, and support contact, then train a baseline gradient-boosted model.
pending
Score current accounts and rank them by predicted churn probability.
pending
Use SHAP to surface the top drivers per account and write the at-risk list back to the database.
pending
Sample output
json
// Sample output
// (generated when the pipeline finishes)
Train a baseline churn model, score active accounts, and return a ranked at-risk list with the top driver per account.
Unlock the rest
The full agent definition, install snippet, and starter task are gated for community members.
Members get the full `.md` agent file, the npm / pnpm install one-liners, a starter prompt that we've tuned against real runs, and the open-source repo when this automation ships there. One email, magic link, done.