Skip to content

Connecting Researchers to Research Grants

A Case Study by AurumNova

Background

The business wants to help its customers (universities) plan which research areas to invest in and make it easier for its faculty staff to win funding for their research.

AdobeStock_302868001

The Challenge

Risky Investments

Deans want to know if they should invest in a growing area of research. However, to ascertain this, they must categorize thousands of grants for each research topic of interest. This can take weeks to months of work each year, and tabulation of the amounts is error-prone.

Experts Are Scarce

Typically, they employ interns to do the above, but they must be trained and frequently lack the experience of deans and researchers. The process involves numerous different parties with differing opinions and areas of expertise. This results in inconsistent categorization of research grants.

Time Wasted

A researcher's area of expertise is niche. However, funding opportunities for research are only broadly categorized. Given that there can be thousands of grants in their area of expertise, researchers spend days to weeks every year searching for relevant funding.

9-Figure1-1-1
Research topics map (Credit: R. Klavans, K. Boyack)

Our Results

  • Leadership can make more informed decisions about which research areas to invest in and which positions to fill.
  • University staff experience increased productivity as they can focus more on their actual research.

These positive outcomes are central to AurumNova's mission. Contact us to discover how AI can significantly enhance your team's capabilities.

 

How We Did It

Divide & Conquer

We build on the work of R. Klavans and K. Boyack, whose method divides all published scientific research into thousands of topics, also known as areas of research.

Mixture of Experts

We trained multiple Large Language Models on millions of published papers, with each model specializing in various research topics.

Validate with Unis

Deans and senior faculty from several top US universities worked with us to validate the accuracy of AI in categorizing grants. We tested a diverse range of representative topics, from hypersonics to urban health.