Radiant Intel

About Radiant Intel

A source is the beginning.
Not the conclusion.

A policy announcement is not an investment thesis. A procurement notice is not a market forecast. The work is connecting evidence to its implications—and making the reasoning open to scrutiny.

Radiant Intel is an AI research platform for investment, policy, and strategy teams. We turn questions about markets, policy, and geopolitics into cited research your team can examine, challenge, and use.

Why we build

The question is what it changes for you.

An investment team needs to test a thesis. A policy team needs to distinguish a proposal from a binding rule. A strategy team needs to understand which markets, suppliers, or counterparties are exposed. Each starts with a different decision—not a request for more information.

Radiant brings source research and structured analysis into a repeatable workflow, so the question, the findings, and the supporting evidence stay connected. The aim is to help your team spend its attention on the implications and the judgment that follows.

In this public China briefing, growth-support measures also introduce conditions businesses need to examine. The useful finding is the relationship between the opportunity and those conditions.

Read the analysis and its limits
Public sample reportFebruary 3, 2026
China Growth & operating environment

Growth: China Brief

What the sources establish

Beijing published a 2026 pipeline of 300 key projects, alongside a national consumption-promotion plan with monitoring and safety requirements.

Why it matters

Growth support and compliance pressure are moving together. Evaluating the opportunity also means examining the conditions attached to it.

Historical sample · Editorial excerptInspect the sources

What that means in practice

Useful research must stand up to questions.

Keep the original evidence

Original-language material is preserved separately from translations and presentation. You need to be able to examine what a source actually says.

Make the reasoning inspectable

Reports bring findings, implications, and supporting citations together. A document's statement and the interpretation built on it are different things.

Keep uncertainty visible

Source coverage has limits. Evidence can be incomplete. Qualifications and gaps belong in the assessment, so you can decide how much weight to give it.

The people behind Radiant

Research experience. Product discipline.

Our team brings together policy research, intelligence tradecraft, product design, and engineering. These are the disciplines needed to make complex research useful in day-to-day work.

Founders & team

MF

Michael Frank

Co-Founder & CEO

Michael brings experience in geopolitical analysis and technology policy to Radiant Intel. Previously a Senior Fellow at the Center for Strategic and International Studies, he focused on AI and emerging technology policy. Before CSIS, he led the Economist Intelligence Unit's Asia technology and geopolitics policy research. He holds a master's in public policy from the University of Chicago.

Philip Rau, Lead Product Designer and Design Engineer at Radiant Intel

Philip Rau

Lead Product Designer and Design Engineer

Phil leads Radiant Intel's product experience and builds the interfaces that deliver it. His work focuses on making complex information usable, drawing on fifteen years designing finance and customer-experience tools. His portfolio includes research and trading products used by Merrill, Morgan Stanley, and RBC.

2430

2430 Group

Co-Founders

2430 Group brings intelligence tradecraft and experience investigating threats to innovation. Its team includes former intelligence professionals, military veterans, technology specialists, and academics. The group works with private industry, government, and academia on identifying foreign espionage risks while preserving productive international collaboration.

Advisors

DK

Dimi Kellari

Strategic Advisor

Dimi advises Radiant Intel on building technology products and bringing them to market. His experience includes leading product and engineering at Cavnue, product and strategy work at Alphabet's Moonshot Factory (X), and investing at Sidewalk Infrastructure Partners. He holds master's degrees from MIT and Imperial College London.

RB

Robin Bew

Advisor

Robin brings experience leading economic forecasting and commercial research. At the Economist Intelligence Unit, he served as Chief Economist, Editorial Director, and Managing Director, overseeing global forecasting, credit-risk, and custom-research teams. Earlier in his career, he was a macroeconomist at the UK Treasury.

RV

Ryan VanGorder

Advisor

Ryan brings experience in investment platforms, private markets, and financial operations. His career spans Quellos, BlackRock, and Opto Investments. At BlackRock, he served as a Managing Director leading finance for the Asia-Pacific region.

BL

Bobby LeGaye

Technical Advisor

Bobby brings software engineering and applied R&D experience, including more than fifteen years building technology in Nike's Innovation NXT and Space Kitchen labs. He holds a degree in computer science from Colgate University.

JM

John Marrett

Macroeconomic Advisor

John brings macro-financial risk expertise focused on Asian economies and politics. He was Director of Financial Risk at the Economist Intelligence Unit and held analytical roles covering East and Southeast Asia for nearly a decade. His experience also includes European Commission business facilitation projects in China and Asia Foundation initiatives in Mongolia.

Programs & affiliations

External programs supporting Radiant Intel's research and product development.

NVIDIA logo

Accepted into

NVIDIA Inception

NVIDIA AI startup program

About the program
Microsoft for Startups wordmark

Joined

Microsoft for Startups

Microsoft startup program

About the program

Judge the work for yourself

Bring a question your team needs to answer.

Explore a complete public briefing, or talk with us about your research scope, source requirements, and the decisions the work needs to support.