Centram Centre for Applied Artificial Intelligence

AI and Machine Learning course for Project Management

Course title
AI & Machine Learning for Project Management

Target audience

 Project Managers, PMO staff, schedulers, Program Directors,
 Data Analysts supporting projects, project portfolio managers
 IT/engineering leads integrating ML into project tools.

Prerequisites
1. Basic statistics and Excel; familiarity with project management concepts (schedule, resources, cost, risks)

2. Recommended: basic Python or R for hands‑on labs (or plan a short pre‑course refresher)
Course learning objectives

By the end of the course participants will be able to:
1. Identify high‑value AI/ML use cases across project lifecycle and portfolio mgt.
2. Prepare and clean project data (schedules, timesheets, cost records, issue logs).
3. Build predictive models for schedule delays, cost overruns, and risk escalation.
4. Use time‑series and probabilistic forecasting for milestones and cash flow.
5. Apply NLP to extract risks/requirements from documents and meeting notes.
6. Use optimization and simulation to support resource levelling, schedule and what‑if analysis.
7. Integrate ML outputs into dashboards and PMIS workflows and monitor model performance.
8. Understand ethical, governance and explainability requirements for project decisioning.
Course format & time commitment

4weeks Boot camp OR 2weeks of physical training and 3 week online study
Delivery: in‑person or online (Zoom + GitHub/Colab + PM tool integrations).

Course Outline

Introduction: AI opportunities in project management

  1. Topics: typical PM problems (schedule slips, budget overruns, resource conflicts, scope creep); where ML adds value vs where PM process change is needed.
  2. Lab: Map ML use cases to your organization; prioritize by impact and data readiness.
  3. Assignment: 1‑page use‑case & data inventory for a chosen project.

Data for projects: sources, quality, and integration

  1. Topics: schedule files (MS Project, Primavera), timesheets, ERP cost data, issue trackers (Jira/TFS), email/meeting notes, sensors/IoT on construction sites; master data, timestamps, event logs.
  2. Lab: Ingest and harmonize schedule + timesheet + cost CSVs; handle missing data & inconsistent units.
  3. Assignment: Data readiness checklist and gap remediation plan.

Exploratory analysis, KPIs, and baseline diagnostics

  1. Topics: EDA for project data, visualizing earned value (EV, PV, AC), trend analysis, lead/lag on dependencies, KPI definitions (CPI, SPI, TOC).
  2. Lab: EDA notebook showing patterns related to overruns and late tasks; baseline rules and thresholds.
  3. Assignment: EDA report highlighting top predictors of delay in sample project.

Predictive modelling: schedule delay & cost overrun classification

  1. Topics: supervised classification/regression, feature engineering for events and resources, train/test split for projects (time‑aware), evaluation metrics (precision/recall, ROC, MAE).
  2. Lab: Build a delay/overrun classifier (logistic regression and tree models) to predict tasks or projects at risk.
  3. Assignment: Model report: features, performance, confusion matrix, business implications.

Time series & probabilistic forecasting for milestones & cash flow

  1. Topics: forecasting methods (ARIMA, Prophet, exponential smoothing), probabilistic forecasts, Monte Carlo schedule risk analysis, PERT vs empirical distributions.
  2. Lab: Forecast milestone completion dates and project burn rate; run Monte Carlo to produce probability of on‑time completion.
  3. Assignment: Forecast with uncertainty intervals and recommended contingency buffer.

NLP for project documents, tickets & stakeholder sentiment

  1. Topics: text pre-processing, classification, topic modeling, entity extraction, meeting‑minute summarization, speech‑to‑text for meetings, sentiment and stakeholder feedback analytics.
  2. Lab: Use NLP to extract risks, action items and themes from meeting notes and issue tracker text; build a classifier to triage ticket priority.
  3. Assignment: Prototype pipeline to extract 5 key risk items and demonstrate triage accuracy.

Resource optimization & scheduling (OR + ML hybrid)

  1. Topics: resource leveling, constrained scheduling, integer programming basics, heuristics, metaheuristics, integrating ML demand forecasts into optimization.
  2. Lab: Formulate resource leveling as an optimization problem (PuLP/OR‑Tools) and compare to rule‑based allocation; use predictive demand for resource planning.
  3. Assignment: Produce an optimized short‑term resource plan and quantify improvement vs baseline.

Simulation & reinforcement learning for planning & execution

  1. Topics: discrete event simulation for site/logistics, scenario planning, RL applications for scheduling/picking/prioritization; when RL is appropriate.
  2. Lab: Simulate task execution variability and test scheduling policies; simple RL agent to prioritize task sequencing under stochastic delays.
  3. Assignment: Compare simulation/RL policy vs heuristics on KPIs (makes pan, lateness).

Deployment, dashboards & PMIS integration

  1. Topics: model deployment patterns, API integration with Jira/MS Project/ERP, dashboarding (PowerBI/Tableau/Streamlit), alerts & escalation workflows, model retraining & monitoring.
  2. Lab: Build a demo dashboard that surfaces at‑risk tasks/projects and model explanations; connect to sample Jira API or synthetic project data.
  3. Assignment: Implementation brief describing how models will be integrated, user roles, SLAs and monitoring plan.

Ethics, governance, change management & capstone presentations

  1. Topics: explainability for decision accountability, privacy (meeting transcripts, personal performance), human‑in‑the‑loop workflows, avoiding automation bias, governance for predictive PM tools.
  2. Capstone: team presentations of end‑to‑end project: problem, data, model, evaluation, deployment and governance plan.
  3. Assessment: peer feedback + instructor scoring.

    Hands‑on labs & tools

  1.  Languages: Python recommended — pandas, scikit‑learn, XGBoost/LightGBM, statsmodels, Prophet, SHAP
  2.  NLP & speech: spaCy, Hugging Face transformers, Speech Recognition, Google/IBM/AWS speech‑to‑text (optional)
  3. Optimization & simulation: Google OR‑Tools, PuLP, SimPy
  4. Dashboards & APIs: Streamlit, Dash, FastAPI, PowerBI/Tableau; project tools APIs (Jira, MS Project, Primavera Web)
  5. Environments: Google Colab, Jupyter notebooks, GitHub
  6. Integrations: sample connectors to Jira/GitHub issues, MS Project exports, ERP/finance CSVs

Datasets & examples

  1. Synthetic project datasets (schedules, timesheets, costs) — create sanitized examples
  2. Public datasets: construction project datasets (where available), software project issue trackers (GitHub/Jira public projects), government procurement/project portals
  3. Internal project archives are ideal for capstones (recommend anonymization)

Assessments

  1. Weekly assignments/labs (50%): EDA, predictive models, optimization, NLP pipelines
  2. Capstone project (35%): group end‑to‑end project, code + report + presentation
  3. Participation & quizzes (15%): readings, demos, in‑class exercises
  4. Rubrics: clarity of problem framing, KPI alignment, data handling, evaluation vs business metrics, explainability & governance plan, reproducibility

      Capstone project ideas

  1. Predicting project completion risk for a portfolio and proposing prioritized mitigations.
  2. Early detection of scope creep from change requests and meeting notes using NLP.
  3. Automated triage of issue/ticket backlog to reduce mean time to resolution.
  4. Resource leveling optimization for multi‑project shared resource pool.
  5. Invoice & cost anomaly detection to flag potential overspend or fraud.

Speech‑to‑text + NLP to produce action‑item tracker from weekly stand-ups.