Hugo Academic CV Theme

Philippe Heitzmann

Technical AI Product Manager

Harvard Rock Center for Entrepreneurship

prev. PM at Arcules Cloud & Data Scientist at J.P. Morgan

Professional Summary

I’m an experienced operations & analytics leader and a recent graduate of Harvard Business School (2025). I previously worked as an operations & analytics lead at Harvard’s Rock Center for Entrepreneurship, as a technical product manager at Arcules Cloud (100 person scale-up) and as a data scientist at J.P.Morgan. In my free time I enjoy pursuing entrepreneurial projects and have built three software apps 0→1, each bootstrapped and revenue-generating. I’m currently seeking BizOps or GTM roles at high-growth companies.

Interests

BizOps
Go-to-Market
Artificial Intelligence
Data Science

Education

Master of Business Administration (MBA), Class of 2025
Harvard Business School
B.S. in Economics, B.A. in International Relations, Certificate in Data Science
Duke University

Apps I’ve built

Building StartupLeads.ai: Shipping a Micro-SaaS That Delivers Same-Week Fundraising Leads to Agencies

Product-Management

Building StartupLeads.ai: Shipping a Micro-SaaS That Delivers Same-Week Fundraising Leads to Agencies

This post walks through how I built StartupLeads through market research, customer interviews, MVP scoping & PRD, technical architecture, pricing tests, building growth loops and early results.

Sep 7, 2025

From Problem Discovery to MVP: How I Built BoostMyRank as a Micro-SaaS for Backlinks

Product-Management

From Problem Discovery to MVP: How I Built BoostMyRank as a Micro-SaaS for Backlinks

This post walks through how I built BoostMyRank through market research, 10 customer interviews, pricing validation at $249/mo, competitive analysis, MVP scoping & PRD, technical architecture, and a one-week build to launch an AI-assisted backlink service supported by an affiliate program.

Sep 7, 2025

Data Science Projects

Predicting Lending Club Loan Defaults with Machine Learning

Machine Learning | Python

Predicting Lending Club Loan Defaults with Machine Learning

Build and deploy ML models to predict Lending Club loan defaults and optimize a high-IRR portfolio, with interactive EDA, Flask API, and Dash apps.

Jan 15, 2022

Predicting House Prices with Stacked ML (Ames Housing)

Machine Learning | Python

Predicting House Prices with Stacked ML (Ames Housing)

Build stacked models on the Ames Housing dataset to predict SalePrice with rigorous cleaning, feature engineering, RFE, and stacking.

Jan 8, 2022

WSJ Sentiment, Reader Engagement, and SP500 Moves

Web Scraping | Python

WSJ Sentiment, Reader Engagement, and SP500 Moves

Scrape 22k WSJ articles, run VADER/TextBlob sentiment analysis, and test relationships with reader comments and S\&P 500 returns. Includes an R Shiny app.

Sep 21, 2020

Visualizing U.S. COVID-19 Cases & Deaths with R Shiny

Data Analysis | Python

Visualizing U.S. COVID-19 Cases & Deaths with R Shiny

An interactive Shiny app to explore NYT COVID-19 case and death data by state, county, and date using bar charts, scatter plots, lollipop charts, and maps.

Jun 1, 2020

Unsupervised ML Projects

Building customer personas for targeted marketing campaigns

Customer Segmentation | LTV:CAC | Python

Building customer personas for targeted marketing campaigns

We grouped 8,950 credit‑card customers into clear, actionable personas based on real spending and payment behavior. The write‑up explains the data we used, how we formed the groups at a high level, what makes each persona distinct (e.g., cash‑advance heavy vs. everyday spenders), and how teams can activate them to tailor offers, credit limits, and messaging.

Mar 15, 2022

Topic Modeling News Headlines to Classify Articles

Natural Language Processing | Python

Topic Modeling News Headlines to Classify Articles

Unsupervised topic modeling on 1.1M ABC News headlines with LDA, LSA, LSI, and HDP; compare scikit‑learn vs Gensim/NLTK preprocessing and visualize separability with t‑SNE.

Feb 10, 2022

Computer Vision Projects

Credit Card Optical Character Recognition with OpenCV + Tesseract

Optical Character Recognition

Credit Card Optical Character Recognition with OpenCV + Tesseract

Built a lightweight, sub-0.5s/image pipeline to read 16-digit card numbers and cardholder names from card photos using OpenCV (template matching) and Tesseract (OCR). On a small, hand-labeled 23-image set, the baseline achieves 48% recall on PAN and 65% on name. With feasibility of low-latency data capture (fraud/risk, checkout autofill) established, next steps include dataset scaling and custom OCR training to reach production readiness.

Sep 7, 2025

Neural Style Transfer: From-Scratch (VGG-19) vs. Pre-Trained (TF Hub)

Neural Style Transfer: From-Scratch (VGG-19) vs. Pre-Trained (TF Hub)

A practical, side-by-side walkthrough of two ways to stylize images: a custom VGG-19 approach you can fine-tune for unique brand looks, and a pre-trained TensorFlow Hub model that ships fast and scales easily. Includes links, code snippets, and the trade-offs product teams care about (control vs. speed, quality vs. effort).

Sep 7, 2025

Road Damage Detection at Scale: YOLOv5 Ensemble for Real-Time, Low-Cost Road Infrastructure Monitoring

Computer-Vision

Road Damage Detection at Scale: YOLOv5 Ensemble for Real-Time, Low-Cost Road Infrastructure Monitoring

Built and benchmarked a smartphone-first road-damage detector using YOLOv5 (with ensembling + test-time augmentation) and Faster R-CNN on the Global Road Damage Detection dataset. Achieved a top-5 leaderboard result (F1 0.68) across 121 teams while meeting a 0.5s/image inference target—enabling practical, low-cost deployment from dashboard-mounted phones. Includes a mapping concept (GPS → segment scores) to guide maintenance prioritization for DOTs and municipalities.

Sep 7, 2025

Smart Video Analytics: Motion Detection for Security & Monitoring

Computer-Vision

Smart Video Analytics: Motion Detection for Security & Monitoring

Led development of an intelligent video monitoring system that automatically detects moving objects in security footage. Successfully evaluated 8 different detection algorithms across 53 test videos, achieving 96% accuracy in ideal conditions and 82% in challenging lighting scenarios. This solution enables automated security monitoring, parking occupancy tracking, and retail analytics without requiring constant human oversight.

Oct 24, 2023

Publications

Philippe Heitzmann (2022). A Computer Vision-assisted Approach to Automated Real-Time Road Infrastructure Management. In arXiv’s Computer Vision and Pattern Recognition.

Code Dataset PDF