Centram Centre for Applied Artificial Intelligence

AI ANALYTICS AND AUTOMATION COURSE FOR PROGRAM MANAGEMENT

Course overview

This course AI‑Assisted Analytics & Automation for Development Program Management is designed for Program Managers, Product Leads, Delivery M PMO staff and engineering managers who want to use AI and automation to improve planning, forecasting, delivery and governance.

Course title

 AI‑Assisted Analytics & Automation for Development Program Management
Target audience

Program Coordinators, Portfolio Managers, Delivery Leads, Product managers, PMO staff, Engineering managers, release/train engineers, DevOps leads.

Course learning objectives

By course end participants will be able to:

  1. Foundations: Metrics, Outcomes & Data Strategy
  2. Topics: Outcome-driven metrics, OKRs vs outputs, leading vs lagging indicators, telemetry sources (VCS, CI/CD, issue trackers, observability), data governance
  3. Lab: Map existing data sources to a metrics taxonomy for a sample program.

Data Engineering for PM Analytics

  • Topics: ETL patterns, event collection (webhooks), data models for issues/commits/builds/releases, data quality, schema design
  • Lab: Ingest sample Jira/GitHub/CI data into a simple analytics DB or DataFrame

Descriptive & Diagnostic Analytics for Programs

  1. Topics: Dashboards, cohort analysis, cycle time, flow metrics, burnup/burndown, root cause analysis, anomaly detection basics
  2. Lab: Build a dashboard (Looker/Tableau/Power BI/Metabase) showing sprint health and cycle time breakdown

 Predictive Analytics: Forecasting & Risk Scoring

  1. Topics: Velocity forecasting, Monte Carlo schedule simulation, features for delivery prediction (story size, dependencies, team changes), model evaluation
  2.  Lab: Build a simple delivery date forecasting model and run Monte Carlo scenarios

Prioritization & Decision Support

  1. Topics: Value/cost/risk scoring, multi-criteria decision-making, demand shaping with AI-assisted prioritization, simulation of trade-offs
  2. Lab: Create a prioritization model combining business value, risk and effort; produce recommended backlog ordering

Automation for Program Operations

  1. Topics: Automating status reporting, meeting prep, release gating, CI/CD triggers, runbooks and playbooks, RPA patterns for administrative tasks
  2. Lab: Implement an automated weekly status report (pulls from issue tracker, CI, test coverage) and posts to Slack/email

Integrating Human-in-the-Loop & Explainability

  1. Topics: When to automate vs assist, confidence thresholds, override flows, explainable models, communicating uncertainty to stakeholders
  2. Lab: Add human approval workflows and transparent explanations to a predictive alert (e.g., “high risk of delay”) and simulate decision flow

MLOps, Monitoring & Model Governance

  1. Topics: Model versioning, monitoring data drift, retraining triggers, logging decisions for audit, access & security concerns
  2. Lab: Set up simple model monitoring (data drift/accuracy alerts) and a retraining plan

Change Management, ROI & Procurement

  1. Topics: Pilot design, success metrics, cost/benefit, vendor assessment for AI tools, procurement considerations, stakeholder engagement
  2. Lab: Draft a pilot plan and ROI case for rolling out an AI-assisted forecasting tool

Capstone Presentations & Roadmap

  1. Topics: Integrating solutions into program lifecycle, scaling from pilot to production, policy checklist
  2. Lab/Project: Present capstone projects and implementation roadmap for an organizational pilot

Hands-on labs

  1.  Ingest and correlate Jira + Git + CI logs to compute flow metrics
  2. Build a predictive model for delivery dates or sprint completion probability
  3. Monte Carlo simulation of program timelines using historical velocities and dependency graphs
  4. Automate weekly stakeholder reports (email/Slack) with attached insights and highlights
  5. Create alerting rules for risk (e.g., rising defect density, blocked tickets) and automated triage suggestions
  6.  Implement an approval workflow that escalates high-risk items to PMO with model explanation
  7. Prototype a prioritization assistant that scores backlog items by value, cost, risk

Capstone project ideas

  1.  End-to-end AI-assisted delivery forecasting system (ingest → model → dashboard → alerts)
  2.  Automated release gating and rollback recommendation engine tied to CI signals and canary metrics
  3.  Backlog prioritization assistant integrated with Jira that suggests sprint scope and expected delivery probability
  4.  Automation to reduce meeting overhead: automated meeting agendas, pre-populated status, action item tracking and follow-ups
  5.  Risk heatmap and dependency analysis across multiple teams with proactive mitigation recommendations

Tools and technologies (recommended)

  1. Data sources: Jira, GitHub/GitLab, Bitbucket, Jenkins/CircleCI/GitHub Actions, Test frameworks, Observability logs (Prometheus, Datadog)
  2.  Data stack: PostgreSQL, BigQuery, Snowflake or simple Pandas + CSVs for labs
  3.  ETL/orchestration: Airflow, Prefect, simple Python scripts, webhooks
  4.  ML & analytics: Python (pandas, scikit-learn, Prophet/ARIMA, PyMC3 for uncertainty), statsmodels, lightGBM/XGBoost
  5. Dashboards/BI: Looker, Tableau, Power BI, Metabase, Grafana
  6.  Automation & integrations: Zapier/Make, GitHub Actions, Selenium/RPA tools, Slack/Microsoft Teams APIs
  7. MLOps/Monitoring: MLflow, Seldon, Prometheus for metrics, Datadog
  8. Collaboration: Confluence, Notion, Slack/MS Teams, Jira

Governance, ethics & risk controls

  1. Human oversight and approval points for automated decisions
  2. Audit logs for model outputs and automated actions
  3. Data privacy and access controls (sensitive HR or personnel data)
  4. Bias and fairness checklists (avoid penalizing teams on raw velocity)
  5. Rollback and fail-safe mechanisms for automation

Deliverables participants will produce

  1. Data mapping and metrics taxonomy for a program
  2. One predictive model or automation workflow with evaluation
  3. Dashboard or report prototype demonstrating actionable insights
  4. Pilot plan + ROI and governance checklist
  5. Capstone presentation and implementation roadmap

Recommended readings & resources

  1. “Accelerate” by Forsgren, Humble & Kim (software delivery metrics)
  2.  Research on Monte Carlo forecasting for software delivery
  3. Articles on flow metrics (flow efficiency, lead time, throughput)
  4. Vendor docs: Jira API, GitHub/GitLab analytics, CI provider docs

Intro M