Methodology | AI Transformation Index
Learn about the Ciridae AI scoring methodology. Understand how durability and opportunity scores are calculated for portfolio companies.
The Ciridae AI Score
Through our work advising on and transforming hundreds of PE-backed companies across industrial, healthcare, professional services, insurance, software, and beyond, we've seen firsthand how AI reshapes competitive dynamics differently in every industry. The patterns we observed led us to build the Ciridae AI Score - a framework that distills two forces determining which portfolio companies will create the most value in the AI era.
Durability
Will this company's market position and profit pool survive AI disruption? We look at the structural advantages that protect value as AI reshapes competitive dynamics.
Opportunity
How much unrealized AI upside is sitting in this business? We identify where AI adoption can drive step-change improvements in efficiency, growth, and margin expansion.
Why AI Changes Everything
AI is not just another technology cycle. The cost of building software has fallen by orders of magnitude, and the cost of performing cognitive work is falling just as fast. Tasks that once required teams of specialists (analysis, underwriting, compliance review, content creation) can now be done by smaller teams augmented with AI, or in some cases by AI alone. This is restructuring entire industries, expanding margins where AI substitutes human labor, creating massive upside where companies harness it, and shifting competitive advantage toward those who adapt fastest.
For PE investors, the impact of AI isn't theoretical, it's already here. AI is already changing how portfolio companies operate, compete, and create value. The question is not whether AI will affect a given business, but how much and in which direction.
The transformation moment: We are at an inflection point where the companies that move first on AI transformation will compound their advantage. The gap between AI-enabled and AI-lagging portfolio companies will widen rapidly across operational efficiency, margin structure, and ultimately enterprise value at exit.
Core Architecture
AI Score = √(Durability × Opportunity)
Geometric mean - rewards companies that are strong on both dimensions
Durability
How durable is the company's market position against AI disruption?
Opportunity
How much unrealized AI step-change exists for this business?
AI Score
Single headline score combining both via geometric mean.
Rating Scale
Each sub-component is scored 1-10, where higher scores indicate a more favorable position for the company. The same scale is used across both Durability and Opportunity.
1-2
Structurally Unfavorable
3-4
Unfavorable
5-6
Mixed / Neutral
7-8
Favorable
9-10
Structurally Favorable
Schumpeter's Return
Joseph Schumpeter described capitalism as a process of creative destruction. New technologies don't just improve existing businesses; they dismantle the economic structures those businesses depend on. For most of the last thirty years, the internet threatened distribution but left production economics largely intact. AI is different. It attacks the production function itself.
When the cost of cognitive work falls by an order of magnitude, the consequences cascade. Entire categories of economic output (analysis, advice, underwriting, compliance, content, coordination) shift from labor-intensive services to near-zero-marginal-cost software. Margins that were protected by expertise evaporate. Intermediaries whose value was "knowing things" get bypassed by systems that know things faster and cheaper. Profit pools migrate from the people doing the work to the platforms and data assets that make the work unnecessary.
This is Schumpeter's gale, arriving all at once. The framework below maps the five mechanisms through which AI restructures competitive dynamics, not as abstract theory but as concrete forces we score for every company we evaluate.
Five Mechanisms of AI Disruption
Substitution
AI replicates the paid outcome (e.g., content, analysis, coding, design, triage).
Disintermediation
AI-enabled platforms/agents bypass intermediaries and "middle layers."
Commoditization
AI lowers entry barriers, increasing competition and compressing margins.
Profit-Pool Migration
Value shifts to adjacent layers (system-of-record, data owners, distribution owners, platforms).
Trust & Regulatory Constraints
Rules or safety requirements limit AI substitution, preserving incumbents.
Why this matters for scoring: Every Durability sub-component maps directly to one of these mechanisms. D1 measures substitution risk. D2 measures profit-pool migration. D3 measures structural barriers that slow these forces. D4 measures disintermediation exposure. D5 measures whether the moat survives once all five forces are in motion.
Durability (0-100)
Durability measures the resilience of the company's end-market economics and its value-chain role to AI-driven changes in substitution, disintermediation, competition, and profit-pool migration.
D = 0.20 × (D1 + D2 + D3 + D4 + D5)
Opportunity (0-100)
Opportunity measures how much unrealized AI step-change exists for this business. If they did AI transformation properly, how much would it benefit them?
This is deliberately separated from Durability. A legal process outsourcer could be massively transformed by AI (that's high Opportunity). Whether competitors also transform and the market compresses is a separate question (that's Durability). Opportunity measures how big AI is for you. Durability measures whether you survive AI.
O = 0.25 × (O1 + O2 + O3 + O4)
The Step-Change Test
The key question: "What's the unrealized step-change?" Companies with large cognitive workforces doing manual knowledge work (claims adjusters, underwriters, analysts) score highest - AI could compress thousands of person-hours into minutes. Companies that are already AI-native have captured much of the step-change already; score what's still ahead, not what's been realized.
The Self-Sufficiency Discount
Companies that are themselves AI-native or have world-class engineering talent should have lower Opportunity scores, not because AI isn't relevant, but because the unrealized potential is smaller. An insurance TPA with 2,000 manual claims adjusters scores higher on Opportunity than an AI-native data platform.
The Cognitive Work Spectrum
Not all cognitive work is the same. Rules-based work (applying procedures, checklists) drives O1 and O2 high but O4 should be tempered. Judgment-based work (expert intuition, pattern recognition) drives O4 high, but only when it happens at volume.
Mostly rules-based, high volume
Medical bill review, compliance checking
Judgment-based, high volume
Specialty underwriting, consumer lending
Judgment-based, low volume
M&A advisory, bespoke consulting
Calibration Examples
AI-Native Data Platform
Owns the infrastructure layer (high D). Already an AI company, doesn't need outside help; unrealized step-change is small (low O).
Physical Retail / Storefronts
Stores, supply chains, and in-person services are physical moats (moderate D). Limited cognitive work to compress; not an AI thesis (low O).
Insurance TPA / Medical Review
Regulatory moats and switching costs (moderate D). Massive unrealized opportunity - thousands of reviewers/adjusters doing cognitive work AI can compress.
Defense Tech with Elite Engineering
Maximum structural protection: classified, government, physical (very high D). Already has elite engineering; doesn't need outside help (moderate O).
Specialty Insurance Broker
Distribution lock-in and regulatory moats (moderate-high D). Deep domain intelligence in complex risk placement that AI can amplify at scale (high O).
Accounting / Advisory Firm
Cognitive labor is substitutable (low D). Thousands of professionals doing manual knowledge work with no internal AI capability = massive unrealized opportunity (high O).