Proprietary Research for Institutional Investors Navigating AI-Driven Discontinuity
Decoding Discontinuity provides institutional-grade research on the companies, infrastructure shifts, and value migrations reshaping the AI economy.
Our work is designed for investors, operators, and strategic decision-makers seeking a deeper understanding of how generative and agentic AI is transforming competitive advantage across software, compute, infrastructure, and enterprise systems.
Unlike our public essays and leadership analysis published on Substack, our institutional research delivers proprietary company-level analysis, technical assessments, strategic frameworks, and market intelligence developed for professional use.
What We Cover
Our research focuses on the emerging fault lines of the AI economy, including:
AI infrastructure and compute markets
Frontier model economics
Inference efficiency and cost curves
GPU, custom silicon, and cloud dynamics
AI-native business models
Value migration across software categories
Public and late-stage AI companies
Competitive positioning in foundation models and agentic systems
Recent Research
Cerebras
An institutional teardown of Cerebras ahead of its IPO, examining inference efficiency, competitive positioning, architecture strategy, and the evolving economics of AI compute.
Fermi
A deep analysis of the emerging infrastructure stack and the second-order implications of AI-native compute systems. (Fermi)
MiniMax
An examination of frontier model economics, product positioning, and the strategic implications of Chinese AI model competition. (MiniMax)
Figma
A breakdown of a beautifully engineered SaaS 2.0 company valued for an agentic future it has not yet fully earned. (Figma)
Klarna
Assessing the first AI-powered commerce network IPO. (Klarna)
CoreWeave
An appraisal of the neocloud business that has become a vital part of the agentic scaling story. (CoreWeave)
Chime
A rigorous review of an influential neobank as it prepares for an IPO. (Chime)
Recognition
“Her firm, Decoding Discontinuity, warned in a report just after Fermi’s IPO that its stock was buoyed by investors making a risky bet that its apparent political connections would pay off.” - Washington Post, April 28, 2026.
”She is trying to add a dose of sober financial analysis to the hyped-up world of AI, without pouring much cold water on the technology itself.” - The Information, March 20, 2025
Research Philosophy
We believe the AI era represents a discontinuity comparable to the emergence of the internet or cloud computing, but unfolding at a significantly greater pace.
The winners of this transition will not simply be the companies with the largest models or the most capital. They will be the organizations best positioned around:
inference economics,
distribution,
workflow integration,
proprietary data,
and durable systems of value capture.
Our research is built to identify these shifts before they become consensus.
Institutional Access
Institutional Research access is offered separately from our public Substack membership. Annual institutional includes:
Proprietary research reports
Company teardowns
Strategic briefings
Direct analyst access
Custom research discussions
Access is limited and managed directly. To inquire about Institutional Research, please contact:
Raphaëlle d’Ornano
Founder & CIO, Decoding Discontinuity
raphaelle@decodingdiscontinuity.com
Or
Mathieu Huet
COO, Decoding Discontinuity
mathieu@decodingdiscontinuity.com
For investors and corporate leaders seeking to understand the transformative nature of the discontinuity being caused by generative and agentic AI, these reports and methodologies offer a playbook for evaluating high-risk AI investments, separating speculation from strategy, and defining defensible value and enduring competitive advantages.