Meridian Database

Overview

In a time of escalating geopolitical tensions, some of the most important stories are being lost in the artificial noise: the reality of deep, ongoing collaboration between the world's two largest powers. Thousands of scientists, engineers, entrepreneurs, trade experts, and more, are working on problems that don't care about flags. Meridian's hope is to act as a counterweight to the artificial geopolitical noise by showing the reality of collaboration between some of the most important US and Chinese companies, businesses, organizations, and even government agencies.

The technical stack is as follows:

• Next.js
• TypeScript
• Tailwind CSS

Meridian is centered around its database of US-China partnerships and their details. To provide more context regarding these partnerships to visitors, they are ranked in order of significance, on a scale from 0-100 in total. The "significance rating" is composed of four main factors:

• Longevity: This considers how long the partnership has lasted, and is weighed 10% on the total significance score. Points are awarded for every year of the partnership's existence in order to judge its stability. A maximum score of 10/10 for the longevity rating can only be achieved if the partnership has lasted 25 continous years or longer. We use an accelerating power curve so longevity yields increasing returns over time, giving smaller early boosts and larger rewards for long-lived partnerships.
• Normalized Scale: This is the primary factor in the final significance rating, weighing 50%. We use a logarithmic scale to measure total capital (or research) output generated as a result of the partnership. A log scale means we give points for adding zeros (e.g., $1M to $10M scores the same jump as $10B to $100B), keeping massive and small partnerships on the same readable scoreboard.
• Rate of Progress/Output: Considers the partnership's rate of progress, or, in other words, its efficiency relative to its age. We use impact/years^0.75 for this, as it rewards new and rapidly-scaling partnerships but also cuts older, massive projects some slack so their score doesn't tank just because they've been operating for 25 years.
• Operational Effectiveness: This is a manual, qualitative score inputted by a researcher that considers the competency of each participating entity in the partnership (or of the joint-venture leadership) and the partnership's overall effectiveness in practice. This catches things that cannot be automatically determined by mere numerical data. Partnerships or joint-ventures that have proven to be extremely effective in every meaningful regard and also indicate highly competent leadership will be rewarded accordingly.

In addition to the US-China partnership index, Meridian also features a database of laws and regulations, from both the US and China, governing business and trade between the two countries, as well as an index of some of the most groundbreaking scientific research papers published by or related to a US-China partnership, and finally, a directory of every individual partner entity involved in a US-China partnership.

Because of the nature of Meridian (small, neutral, and unfunded), a set of fine-tuned local LLMs are used to search the web and generate text regarding the respective entities.

Some things are changing quickly, so some details on this page may no longer be completely accurate, but the idea will be the same.

As of June 2026, Meridian's website launched as a public beta, though many features are still WIP and/or unavailable. This is a project created for the common good of humanity, and is not funded.