Certificate 01 · Weeks 1 to 4
Certificate in Python pandas and SQL Foundations
Data engineering foundation. Answer the same question twice, then reconcile.
This section builds the programming and querying foundation for the rest of the bootcamp. Week 1 introduces Python variables, selections, loops, functions, lists, dictionaries and basic debugging. Week 2 is devoted to pandas DataFrames: reading files, inspecting columns, handling missing values, selecting records, combining tables and calculating grouped summaries.
What You Will Be Able to Do
- Translate a short requirement into pseudocode and a small Python function.
- Use selections, loops and collections, and diagnose common logic errors.
- Read, clean, filter, combine and summarize tabular data with pandas.
- Write SQL queries with filters, aggregates and appropriate joins.
- Use an introductory CTE or window function for a defined analytical question.
- Reconcile outputs and explain missing values, duplicates and assumptions.
Live Meetings · 8 Meetings, 16 to 24 Hours
- Week 1 · Meeting one: Introduction to Python: variables, types, conditions and basic collections
- Week 1 · Meeting two: Loops, functions, file handling and debugging a small program
- Week 2 · Meeting one: pandas DataFrames: import, inspection, types and missing values
- Week 2 · Meeting two: Filtering, joins, grouping, derived columns and result checks
- Week 3 · Meeting one: SQL fundamentals: tables, keys, SELECT, filtering and sorting
- Week 3 · Meeting two: Aggregates, GROUP BY, nulls and joining related tables
- Week 4 · Meeting one: SQL for analytics: CTEs, introductory windows and business measures
- Week 4 · Meeting two: Python and SQL project review, reconciliation and certificate assessment
Certificate project. A reproducible pandas notebook and SQL query collection answering a business brief. Submissions include source data, expected result checks, a short data dictionary and a README explaining how to rerun the work.
Entry preparation. Basic computer and file-management skills, arithmetic and introductory algebra. Prior Python or SQL is not required. Complete the environment setup and short programming readiness exercises before the first class.
PythonpandasSQLCTEs & WindowsReconciliation
Certificate 02 · Weeks 5 to 8
Certificate in Data Modeling and ETL
Data engineering development. Turn separate files into a controlled workflow.
This section turns separate files and application data into a controlled preparation workflow. Learners design small relational models, identify keys and table grain, and distinguish transactional structures from analytical structures. They ingest CSV and JSON files and practice reading an application programming interface, or API, using a supplied classroom service and saved responses.
What You Will Be Able to Do
- Define table grain, keys, relationships and a source-to-target mapping.
- Ingest files and a paginated API response using Python.
- Apply consistent data types, cleaning rules and duplicate handling.
- Implement a small incremental load with documented update rules.
- Test completeness, validity, uniqueness and referential integrity.
- Document exceptions, lineage, access needs and safe rerun behavior.
Live Meetings · 8 Meetings, 16 to 24 Hours
- Week 5 · Meeting one: Database design: grain, keys, relationships and normalization
- Week 5 · Meeting two: Analytical modeling, fact and dimension concepts and source mapping
- Week 6 · Meeting one: File ingestion: CSV, JSON, schemas and error handling
- Week 6 · Meeting two: API ingestion: pagination, credentials, retries and saved responses
- Week 7 · Meeting one: ETL and ELT: staging, transformations and repeatable loading
- Week 7 · Meeting two: Incremental updates, duplicates, late records and rerun behavior
- Week 8 · Meeting one: Data quality tests, lineage, ownership and access rules
- Week 8 · Meeting two: Pipeline demonstration, exception review and certificate assessment
Certificate project. A repeatable Python and SQL workflow that ingests multiple sources, loads modeled tables and generates a quality report. Evidence includes a schema diagram, versioned transformations, a rejected-record file, reconciliation totals and a rerun test.
Entry preparation. The first certificate or equivalent Python, pandas and SQL skills. Individual enrollees complete a readiness task involving a join, a grouped summary and a small Python transformation, then receive the project brief and starting data.
Data ModelingETL & ELTAPI IngestionData QualityGit
Certificate 03 · Weeks 9 to 12
Certificate in Data Warehousing and Cloud Pipelines
Data engineering capstone. Make the refresh repeatable, then hand it off.
This section organizes prepared data into a small analytical warehouse and makes its refresh repeatable. Learners build fact and dimension models and use dbt to organize SQL transformations, tests and documentation. Workflow orchestration introduces task dependencies, schedules, retries, logs and recovery after a failed step.
What You Will Be Able to Do
- Build a small analytical model with defined facts, dimensions and grain.
- Organize transformations, tests and documentation in a dbt project.
- Define an orchestrated workflow with dependencies and failure handling.
- Use a classroom storage environment with scoped access and cost controls.
- Compare a local transformation with a guided distributed-processing example.
- Demonstrate a tested refresh and hand off an analysis-ready dataset.
Live Meetings · 8 Meetings, 16 to 24 Hours
- Week 9 · Meeting one: Warehouse structure and dbt transformation models
- Week 9 · Meeting two: dbt tests, documentation, model dependencies and change tracking
- Week 10 · Meeting one: Orchestration: tasks, dependencies and schedules
- Week 10 · Meeting two: Retries, logs, failure investigation and a recovery exercise
- Week 11 · Meeting one: Cloud object storage, columnar files, permissions and cost awareness
- Week 11 · Meeting two: Partitioning, scaling tradeoffs and a guided distributed-processing lab
- Week 12 · Meeting one: Engineering capstone integration and quality review
- Week 12 · Meeting two: Refresh demonstration, certificate assessment and data handoff
Certificate project. An engineering capstone containing ingestion code, modeled warehouse tables, dbt tests and documentation, an orchestrated refresh, and a runbook. The demonstration includes a new data delivery, a failed quality check and a corrected rerun. The handoff identifies table grain, freshness expectations and approved uses.
Entry preparation. The first two certificates or equivalent skills in Python, SQL, relational modeling, ETL and quality testing. Individual enrollees receive a prepared ingestion baseline but must explain and modify it. Classroom cloud access and spending limits are established before this section.
Warehouse ModelingdbtOrchestrationCloud StorageRunbook
Certificate 04 · Weeks 13 to 16
Certificate in Applied Statistics and Regression
Data science foundation. Describe uncertainty before you predict.
This section uses the engineering handoff to frame analytical questions and examine patterns before modeling. Learners perform exploratory data analysis, calculate descriptive measures and create clear visualizations. Probability, sampling, confidence intervals and introductory hypothesis testing help learners describe uncertainty and distinguish an observed association from a causal explanation.
What You Will Be Able to Do
- Turn a business question into a target, decision context and evaluation plan.
- Profile data and communicate distributions, relationships and anomalies.
- Interpret basic probability, confidence intervals and hypothesis-test results.
- Choose a split that respects the time or grouping structure of the problem.
- Compare a regression model with a baseline using suitable error measures.
- Explain residual patterns, uncertainty and limits on the conclusions.
Live Meetings · 8 Meetings, 16 to 24 Hours
- Week 13 · Meeting one: Business questions, targets and exploratory data analysis
- Week 13 · Meeting two: Visualizations, missingness, outliers and an initial findings brief
- Week 14 · Meeting one: Probability, sampling and uncertainty
- Week 14 · Meeting two: Confidence intervals, introductory tests and interpretation limits
- Week 15 · Meeting one: Machine learning workflow, baselines and appropriate data splits
- Week 15 · Meeting two: Preprocessing pipelines, cross-validation and leakage prevention
- Week 16 · Meeting one: Regression: linear and tree-based models, errors and residuals
- Week 16 · Meeting two: Regression project demonstration and certificate assessment
Certificate project. An exploratory analysis and regression notebook with a written question, data audit, baseline comparison, validation results and final test evaluation. A concise stakeholder brief describes the practical interpretation, major sources of error and limits of the evidence.
Entry preparation. The first certificate or equivalent Python, pandas and basic SQL skills, together with basic algebra and comfort interpreting graphs. Direct entrants need not complete all engineering sections; they receive a validated analytical dataset and demonstrate readiness in pandas and basic statistical interpretation.
EDAProbabilityHypothesis TestingRegressionBaselines
Certificate 05 · Weeks 17 to 20
Certificate in Classification and Segmentation
Applied machine learning. Choose metrics that reflect the decision.
This section extends the machine learning workflow to classification and unsupervised segmentation. Learners build logistic regression and tree-based classifiers, select metrics that reflect the decision, and examine how classification thresholds affect false positives and false negatives. Feature engineering and limited hyperparameter tuning take place within the training and validation process.
What You Will Be Able to Do
- Define a classification target and compare candidate models with a baseline.
- Interpret a confusion matrix, precision, recall and decision thresholds.
- Build preprocessing and feature steps that avoid target leakage.
- Tune a limited set of model options using appropriate validation data.
- Explain model behavior and inspect errors across relevant subgroups.
- Create, profile and assess segments using an introductory clustering method.
Live Meetings · 8 Meetings, 16 to 24 Hours
- Week 17 · Meeting one: Classification: logistic regression, trees and baselines
- Week 17 · Meeting two: Confusion matrices, precision, recall, imbalance and thresholds
- Week 18 · Meeting one: Feature engineering, encoding and scaling
- Week 18 · Meeting two: Controlled model comparison and limited hyperparameter tuning
- Week 19 · Meeting one: Interpretation, feature importance and error analysis
- Week 19 · Meeting two: Subgroup evaluation, privacy, intended use and model documentation
- Week 20 · Meeting one: Clustering, segment profiles and stability
- Week 20 · Meeting two: Classification and segmentation review and certificate assessment
Certificate project. A classification experiment report and a separate segmentation analysis. The portfolio includes a preprocessing pipeline, baseline and model comparisons, a chosen threshold with justification, segment profiles and a model card describing intended use, evaluation and limitations.
Entry preparation. The applied statistics and regression certificate or equivalent experience with pandas, supervised learning, train-validation-test separation and model evaluation. A readiness exercise checks baseline construction, leakage awareness and interpretation of prediction errors.
ClassificationThresholdsLeakage PreventionClusteringModel Card
Certificate 06 · Weeks 21 to 24
Certificate in Forecasting and Model Delivery
Integrated capstone and handoff. From a dependable pipeline to a usable result.
This section completes the connection between a dependable data pipeline and a usable analytical result. Time-series instruction introduces trend, seasonality, lagged features and chronological evaluation. Learners compare a simple forecast with a naive or seasonal baseline and explain the effect of the forecast horizon on validation.
What You Will Be Able to Do
- Construct a time-aware baseline and evaluate a forecast chronologically.
- Create lagged or rolling features without using future information.
- Package a selected model and its preprocessing for repeatable batch scoring.
- Validate new inputs and document model and data versions.
- Define checks for data drift, scoring failures and later performance review.
- Demonstrate the complete workflow and explain operating and handoff steps.
Live Meetings · 8 Meetings, 16 to 24 Hours
- Week 21 · Meeting one: Time series: trend, seasonality, lagged features and baselines
- Week 21 · Meeting two: Chronological backtesting, forecast error and horizon selection
- Week 22 · Meeting one: Model delivery: saved artifacts, dependencies and batch scoring
- Week 22 · Meeting two: Input checks, logs, drift monitoring and a model operations plan
- Week 23 · Meeting one: Integrated implementation and a new-batch rehearsal
- Week 23 · Meeting two: Technical review, corrections and stakeholder presentation practice
- Week 24 · Meeting one: Capstone demonstrations and individual technical explanations
- Week 24 · Meeting two: Final assessment, portfolio review and technical handoff
Certificate project. A forecast exercise and an integrated capstone that ingests a new data batch, checks it, refreshes the prepared dataset and produces checked predictions or a forecast. Submissions include the repository, evaluation report, model card, scoring instructions, monitoring plan and stakeholder presentation.
Entry preparation. The preceding data science certificates or equivalent modeling and evaluation skills, together with the ability to work with a documented pipeline. Direct entrants receive an engineering baseline; the overall professional certificate still requires all six certificates and the integrated capstone.
Time SeriesBacktestingBatch ScoringMonitoringCapstone
