Data Scientist & ML Engineer · North Carolina

Colby Reichenbach,
Data Scientist & ML Engineer

I design and ship end-to-end ML systems: from data pipelines and model governance to production APIs and deployed frontends. Real data. Deployed code. Measurable outputs.

UNC Logo Portfolio 2026
Colby Reichenbach

Colby Reichenbach

Data Scientist

Use Mouse/Touch to Interact
Contact QR Code
colbyrreichenbach@gmail.com /in/colby-reichenbach /ColbyRReichenbach

Production Work

A collection of dashboards, pipelines, and experiments.

Featured
Pulse Tracker multi-tenant AI coaching SaaS Next.js Supabase — Colby Reichenbach

Pulse Tracker: Multi-Tenant AI Coaching SaaS

A live, multi-tenant AI coaching SaaS built solo. Production guardrails, LLM safety evals, and real users. Not a demo.

View Project
Featured
SPEC-NYC production automated valuation model MLOps dashboard — Colby Reichenbach

SPEC-NYC: Production Automated Valuation Model

1M+ real NYC property transactions. Production MLOps, not a notebook.

View Project
Featured
Proactive Retention Agent MLOps XGBoost churn prediction Streamlit dashboard — Colby Reichenbach

Proactive Retention Agent: MLOps Churn Pipeline

Real telecom data. The system tells retention teams who to call, in what order, and exactly why, before they churn.

View Project
DraftKings Sentinel responsible gaming analytics dbt Snowflake React dashboard — Colby Reichenbach

DraftKings Sentinel: Responsible Gaming Intelligence

Built for one specific DraftKings role. Behavioral risk modeling, state regulatory compliance logic, and an AI-assisted analyst workbench.

View Project
ShelfOps retail inventory intelligence platform FastAPI Kafka XGBoost React TypeScript — Colby Reichenbach

ShelfOps: Retail Inventory Intelligence Platform

Full-stack retail intelligence platform. Forecasting, event streaming, and a multi-level command center built for the way retail teams actually operate.

View Project

I build systems, not demos.

Most data science portfolios stop at notebooks. Mine doesn't.

I'm a UNC Chapel Hill graduate (B.S. Biology, Data Science minor) who has independently shipped production ML systems, full-stack data platforms, and AI-powered applications. The kind of work that typically requires entire teams.

  • SPEC-NYC is a production AVM trained on 1M+ real NYC property transactions with a full MLOps lifecycle: MLflow experiment tracking, champion/challenger model governance, and drift monitoring.
  • ShelfOps is a full-stack inventory intelligence platform with async FastAPI, Celery task queues, Kafka-compatible event streaming, XGBoost/LSTM forecasting, and a React/TypeScript frontend.
  • Pulse Tracker is a live multi-tenant AI coaching SaaS built solo using AI-assisted development, featuring Promptfoo safety evals, Supabase RLS, and Redis rate limiting in production.

I use AI-assisted development as a core part of how I work. It lets me operate across the full stack independently and ship faster than traditional team structures allow. In a field shifting toward AI-augmented workflows, I'm already there.

Education

UNC Chapel Hill Logo
UNC Chapel Hill B.S. Biology, Minor in Data Science
  • Production ML ETL, model training, serving, monitoring, and governance
  • Full-Stack FastAPI backends, React frontends, deployed and running
  • AI-Native Shipping solo with Claude Code and Cursor as force multipliers

Stack

PythonSQLXGBoostMLflowSHAPdbtPandas FastAPIDockerReact/TypeScriptNext.jsRedisOpenAI/Gemini APIs