Forward-deployed engineer & CS infrastructure architect

Peter
Sooter.

I build systems at the intersection of enterprise software and customer outcomes. I don't manage customers. I build the automation that does it, then focus entirely on judgment and relationships. Based in Atlanta, GA. Fluent in English, Portuguese, and Spanish.

leaf-brain on GitHub LinkedIn
16
Scheduled agents
40+
Hours/week recovered
30+
Claude Code skills
15+
Years enterprise CS

What I've built

churn intelligence
customer-health-os

Python implementation of the 7-signal churn scoring model powering leaf-brain. Accepts a CSV of account signals, outputs risk scores, confidence levels, and recommended interventions. Calibrated against real post-mortems.

Python pandas scikit-learn
View on GitHub
trajectory analysis
k-shape-divergence

Account trajectory analysis module. Takes engagement timeseries data and classifies accounts as accelerating toward value or stalling toward churn, within the first 60 days, before divergence becomes irreversible.

Python pandas numpy
View on GitHub
data engineering
automotive-etl-pipeline

Production ETL pipeline for a global automotive manufacturer. Ingests complex daily data dumps from multiple source formats, normalizes and transforms into production-ready analytical datasets. Handles schema inconsistencies, failed delivery recovery, and incremental loading.

Python pandas SQL
View on GitHub
churn intelligence
customer-health-os

Standalone Python implementation of the 7-signal composite churn scoring model. Accepts a CSV of account signals, outputs a ranked risk report with tier classification, top risk drivers, and recommended actions. 21 tests, zero external dependencies.

Python pandas scikit-learn
View on GitHub
trajectory analysis
k-shape-divergence

Account trajectory classifier. Computes stall and acceleration scores independently from engagement timeseries data, classifies accounts as accelerating or stalling within the first 60 days — before divergence becomes irreversible.

Python pandas numpy
View on GitHub
cs operations
meeting-quality-framework

10-dimension meeting quality scoring system with trend tracking, rolling averages, and low-value streak alerts. CLI for interactive scoring and portfolio reporting. Triggers intervention prompts when three consecutive meetings fall below threshold.

Python CLI
View on GitHub

Stack

Languages & Data
Python SQL JavaScript Bash / Shell
AI & Automation
Claude API / Claude Code n8n FastAPI REST APIs / Webhooks
Data & Analytics
Power BI Tableau MongoDB AWS Athena
Infrastructure
Docker Azure macOS launchd Git / GitHub

About

15+ years running enterprise CS and technical operations across SaaS, agtech, and CX platforms. The through-line isn't the industry. It's always been the same: diagnose what's broken operationally, build the infrastructure to fix it, make it run without babysitting.

At Worthix I built the full CS org from zero: 30+ CSMs, the training curriculum, the health scoring system, and the Power BI dashboards that replaced 5 weekly status meetings. At Leaf Agriculture I inherited a manual operation and built leaf-brain. The work is always the same. Figure out what shouldn't require a human, then build the thing that doesn't.

Trilingual: English, Portuguese, Spanish.

location Atlanta, GA
languages EN / PT / ES
focus FDE, RevOps, Technical CS
linkedin peter-sooter