Talk to Me

Shahan Ahmed / Data Products & Systems

Products

From raw data to working systems

A curated list of data products, machine learning pipelines, AI agents, dashboards, and analytics systems built from raw datasets into usable outputs.

Product library

Built systems

4 entries
ML Pipeline

93%

Accuracy · 100% Recall

Cancer Risk Prediction Pipeline

Identifies high-risk cancer patients at the moment of admission.

End-to-end ML pipeline on Databricks that flags high-risk patients from CMS Medicare claims. Delta Lake medallion architecture, MLflow tracking, and 100% recall — no critical cases missed.

PySparkDelta LakeMLflowscikit-learn
AI Agent

64+

Emails Sent

ExForge

Autonomous lead generation — from Google Maps to cold email inbox.

AI agent that scrapes businesses, scores leads with Claude Haiku, audits websites, generates PDF reports, and sends personalized cold emails. Zero manual steps. Runs on a daily schedule.

PythonClaude HaikuPlaywrightSQLite
Analytics

$263K

Identified Savings

Austin Municipal Court Analytics

Data-driven policy interventions for a city court system.

Assessment of 19,906 court cases in Austin covering operational triage, equity monitoring, fiscal impact sizing ($263K savings), and prioritized policy recommendations for city leadership.

PythonPandasSeabornStatistical Modeling
BI Dashboard

100%

Model Accuracy

QSR POS Sales Dashboard

Revenue optimization for a multi-location restaurant chain.

Analyzed 1,743 POS transactions across 10 locations. ML models hit 100% accuracy; revenue analysis surfaced $4–5K in monthly upside. Delivered as a live interactive React dashboard.

ReactPythonRandom ForestRecharts