Free Access · Self-Paced · Industry Datasets

Build Real ML Projects.
Get Certified.

Four industry capstone tracks — Banking, Consulting, FinTech, and Insurance — using production-scale datasets. Run Python in your browser. Earn signed certificates with official credentials.

Explore TracksCertificate Details
4
Industry Tracks
Banking, Consulting, FinTech, Insurance
30+
Project Phases
Across all tracks combined
800K+
Data Points
Real-world production datasets
Free
Track Access
Certificates from $99.99

How It Works

From first click to certified — your path through the program

01

Choose Your Track

Select one of four industry specializations — all completely free to access and work through.

02

Open in Colab or Locally

Open the notebook in Google Colab with one click, or download it to run in your own Jupyter environment.

03

Complete All Phases

Work through 7–9 guided phases: EDA, feature engineering, modeling, and business insights.

04

Get Certified

Purchase a signed PDF certificate for your portfolio and LinkedIn.

Industry Tracks

Each track is a complete end-to-end capstone project modeled after real industry workflows. Choose the sector that aligns with your career goals.

🏦

Banking & Finance

FREE

Credit Scoring Engine — End-to-End ML Project

8–12 hoursIntermediate7 Phases

Predict loan defaults using the Home Credit dataset. Build credit scoring models with Logistic Regression, Random Forest, and XGBoost, then deliver business recommendations for tiered approval systems.

📊
Dataset
Home Credit Default Risk — 300K+ loan applications with demographics, financial history, and bureau scores

What You'll Build

End-to-end credit scoring pipeline
3-model comparison with ROC-AUC benchmarks
Tiered loan approval strategy (auto-approve / review / decline)
Feature importance analysis & risk factor report
PythonPandasScikit-learnXGBoostMatplotlibSeaborn

Project Phases

1
Data Collection
Load real Home Credit application data
2
Preprocessing
Handle missing values, outliers, encode categoricals, normalize
3
EDA
Visualize distributions, correlations, and default rates by segment
4
Feature Engineering
Create debt-to-income, credit utilization, annuity burden, age features
5
Model Building
Train Logistic Regression, Random Forest, and XGBoost classifiers
6
Evaluation
Compare models on Accuracy, ROC-AUC, Precision, Recall
7
Insights
Business recommendations and risk segmentation
Start Free Project8–12 hours · 7 phases · Intermediate
💼

Consulting & Strategy

FREE

AI Strategy Engagement — BakeHouse Global

10–15 hoursAdvanced8 Phases

Conduct a full consulting engagement for a multinational bakery franchise: assess AI readiness, build ROI models, map digital maturity, segment franchises, and deliver an executive strategy deck.

📊
Dataset
BakeHouse Global — 48 franchise locations across US, Japan, Australia with 300+ customers

What You'll Build

Franchise performance scorecard with percentile rankings
5-dimension AI readiness assessment framework
NPV / IRR financial models for 3 AI use cases
4-wave adoption roadmap with stakeholder impact matrix
PythonPandasScikit-learnSciPyMatplotlibSeaborn

Project Phases

1
Client Data Room
Data ingestion & business profiling with Pandas
2
Operational Diagnostics
Franchise performance benchmarking
3
AI Readiness Assessment
Quantitative scoring across 5 dimensions
4
ROI Modeling
NPV, IRR, payback for 3 AI use cases
5
Digital Maturity Map
Maturity heatmap across business units
6
Strategic Segmentation
Franchise clustering for adoption waves (K-Means)
7
Executive Strategy Deck
Charts, roadmap, and recommendations
8
Final Recommendation
Implementation plan & change management
Start Free Project10–15 hours · 8 phases · Advanced
⚡

FinTech & Innovation

FREE

Digital Payments Platform Analytics

10–14 hoursIntermediate9 Phases

Analyze 49K+ payment transactions across 5 digital methods. Build a fraud detection engine, segment customers for neobanking, and create a real-time transaction risk scoring system.

📊
Dataset
WanderBricks — 49,638 transactions, 5 payment methods, 45,958 bookings, 100+ countries

What You'll Build

Unified payment data warehouse from 4 source tables
Fraud detection engine (unsupervised + supervised)
4-persona customer segmentation for neobanking tiers
Transaction risk scoring system with SHAP explanations
PythonPandasScikit-learnXGBoostMatplotlibSeaborn

Project Phases

1
Data Collection
Join payments, bookings, users, and properties
2
Preprocessing
Parse timestamps, handle nulls, derive transaction features
3
Payment Gateway EDA
Transaction volumes, method adoption, success rates
4
Feature Engineering
Transaction velocity, user risk profiles, payment patterns
5
Fraud Detection
Isolation Forest anomaly detection + Gradient Boosting
6
Customer Segmentation
K-Means clustering for neobanking personas
7
Transaction Risk Scoring
4-tier risk classification with XGBoost + SHAP
8
Evaluation
ROC-AUC, precision-recall, cluster profiles, risk tiers
9
Business Insights
FinTech recommendations and deployment architecture
Start Free Project10–14 hours · 9 phases · Intermediate
🛡️

Insurance & Risk

FREE

Healthcare Fraud Detection & Risk Assessment

12–16 hoursAdvanced9 Phases

Detect fraudulent healthcare claims using 409K+ claim lines. Build anomaly detection, supervised fraud classification, and claims cost prediction models for actuarial pricing.

📊
Dataset
HealthVerity — 409,825 claim lines, 1,497 patients, 7,905 providers, 47 U.S. states

What You'll Build

Multi-level feature engineering (provider, patient, claim)
Anomaly-based fraud detection system (Isolation Forest)
Supervised fraud classifier with precision-recall optimization
Actuarial cost prediction model for claims pricing
PythonPandasScikit-learnXGBoostMatplotlibSeaborn

Project Phases

1
Data Collection
Load real HealthVerity claims data from CSV
2
Preprocessing
Parse dates, handle nulls, type casting, deduplication
3
EDA
Claims distributions, geographic patterns, provider analysis
4
Feature Engineering
Build 20+ fraud indicators and risk features
5
Anomaly Detection
Unsupervised fraud detection with Isolation Forest
6
Fraud Classification
Supervised model using XGBoost, Random Forest
7
Claims Cost Prediction
Actuarial pricing model (GLM / Gradient Boosting)
8
Evaluation
ROC-AUC, Precision-Recall, risk segmentation
9
Business Insights
Recommendations for underwriting and claims ops
Start Free Project12–16 hours · 9 phases · Advanced
Certificate of Completion
Banking & Finance
Credit Scoring Engine — End-to-End ML Project
Megersa Daksa (PhD)
Founder & Chief Economist
🏅
Official Credential

Professional Certificates

Earn a signed, verifiable certificate for each track you complete. Add it to LinkedIn, your resume, or portfolio to demonstrate real-world ML expertise.

Official certificate number & credential ID
Signed by Megersa Daksa (PhD) — Founder & Chief Economist
Professional landscape PDF with gold seal
Track-specific title, date, and credential ID
Shareable on LinkedIn & digital portfolios
$99.99per track|$199.99all 4 tracks
Bundle saves $200

Frequently Asked Questions

Everything you need to know about the program

Are the internship tracks free?

Yes. All 4 tracks are completely free to access — including the datasets, code, and in-browser Python environment. Work through real-world ML projects at no cost whatsoever.

What do certificates cost?

Certificates are optional credentials you can purchase after completing all phases of a track. They cost $99.99 per track, or $199.99 for all 4 (saving $200).

Is the program self-paced?

Absolutely. Work through each phase at your own speed. Your progress is saved automatically across sessions. There are no deadlines or cohort schedules.

What's included in each certificate?

A professional landscape PDF featuring an official certificate number, signed by Megersa Daksa (PhD) — Founder & Chief Economist, gold seal, track-specific title, and your completion date.

What tools or setup do I need?

Just a web browser. All projects run directly in your browser using Pyodide (in-browser Python). No Python installation, no Jupyter setup. You can also open any track in Google Colab for free.

How long does each track take?

Tracks range from 8–16 hours depending on the specialization. Banking is 8–12 hours, Consulting 10–15, FinTech 10–14, and Insurance 12–16. These are estimates — take as long as you need.

Can I get a refund on a certificate?

Yes. Contact us within 7 days of your certificate purchase for a full refund, no questions asked.