Executive Summary

This prototype demonstrates how structured portfolio data and generative AI can provide preliminary recommendations to guide strategic application rationalization.

The GenAI App Rationalization Advisor is an AI-assisted prototype designed to help enterprises explore how complex application portfolios could be optimized. Using uploaded CSV data, it provides preliminary guidance on retaining, retiring, modernizing, or consolidating applications and generates a portfolio-level report with AI reasoning.

Overview

Inspired by my experience as an Application Portfolio Manager at GE Vernova, this prototype illustrates the challenges of managing large enterprise portfolios, including redundant or outdated applications and compliance or security risks.

By automating preliminary analysis with AI, it demonstrates potential reductions in manual workload and accelerates insight generation. The recommendations are illustrative and intended to support decision-making exploration rather than replace human judgment.

Key features

  • AI-Assisted Recommendations: Provides preliminary guidance using application metadata and business context.
  • Redundancy & Risk Indicators: Highlights legacy systems, technical redundancies, and potential compliance considerations.
  • Portfolio-Level Report: Generates a CSV/PDF report illustrating recommendations and AI reasoning.
  • Prototype Scalability: Designed to demonstrate handling larger portfolios via CSV inputs.
  • Transparent AI Prompts: Illustrates reasoning and traceability of recommendations.

Target users

  • CIOs, Application Portfolio Managers, and Enterprise Architects (for evaluation/demo purposes)
  • Application Owners
  • Finance and Procurement Teams
  • Security and Compliance Teams

Business value

  • Reduction: Demonstrates identification of redundant or low-value applications.
  • Supports Transformation: Provides illustrative insights for cloud migration, M&A integration, and digital transformation planning.
  • Operational Efficiency: Reduces preliminary manual analysis effort.
  • Strategic Alignment: Improves transparency of potential rationalization decisions.

Technology overview

The prototype integrates Generative AI with structured CSV portfolio data to demonstrate rationalization analysis and reporting.

  • Stack: Python, GPT-3.5-turbo, Tkinter, CSV import
  • Inputs: Application inventory, usage data, cost, compliance risk, and owner input
  • Outputs: Illustrative disposition recommendations with rationale and a portfolio-level report (PDF/CSV)

MVP scope

  • Upload application inventory and usage data via CSV for analysis
  • Generate AI-assisted recommendations for retaining, retiring, modernizing, or consolidating applications
  • Export a portfolio-level report demonstrating recommendations and rationale

Roadmap (Planned)

  • Feedback Loops: Capture user validation on AI-assisted recommendations to improve trust and learning
  • Governance Guardrails: Protect critical applications, require human approval, and show confidence scores to ensure compliance
  • CMDB Integration: Connect with enterprise CMDBs to enable richer, automated portfolio analysis
  • Persona-Specific Prompts: Provide tailored AI guidance for CIOs, Application Owners, Finance, and Security teams
  • Conversational AI Interface: Enable interactive, question-driven exploration of portfolio insights
  • Embedding Similarity & Semantic Search: Detect functionally similar applications even if names differ
  • Build trust, ensure compliance, and unlock deeper insights for smarter rationalization