New AI Startup Proemial Launches Scientific Research Platform

April 30 2024, 00:25
Mads Rydahl, Siri’s first Head of Product and Design, has funneled his extensive knowledge of AI and Large Language Models (LLMs) into a new company. Proemial is a startup, co-founded by serial entrepreneur Geet Khosla and tech veteran Brian Pedersen, with a mission to tackle the most significant problem of accessing credible scientific research and knowledge through search engines and Artificial Intelligence-based tools.
Proemial, supported by well-known tech leaders such as the Amazon CTO, is now launching their free knowledge platform, ‘proem’, which answers user questions backed by scientific research papers and leverages AI to summarize the findings, helping non-experts easily understand and share even complex knowledge.

"Today’s search and AI-powered tools often hallucinate false information, making it look credible, which only adds rocket fuel to the raging fire of misinformation we face globally in every sector of the media, education, policy, and knowledge economy," says Rydahl, who with Proemial wants to give people access to up-to-date, research-backed knowledge. “Proemial is building a platform that makes it possible, for the first time, for anyone to truly understand cutting-edge scientific research; at any level of expertise, and apply the insights in their daily lives; in the classroom, boardroom, research lab, or newsroom."
 
Based in Aarhus, Denmark, Proemial is founded by (from left) tech veteran Brian Pedersen, serial entrepreneur Geet Khosla, and Mads Rydahl, Siri's first Head of Product and Design).
The team built a simple, yet powerful, way to get answers to complex, important questions supported by references from the largest repository of scientific research in the world. Scientific knowledge is constantly evolving and proem will keep users up-to-date on the latest insights by adding over 410,000 new research papers every month. The platform will be open and free, allowing anyone to further dive into specific scientific papers, using LLMs to summarize and explain complex topics in the paper and customize an inspiring feed where users can learn, discuss, and connect with others over new insights and breakthroughs.

Uniquely, proem goes beyond just providing summaries of research papers, by allowing users to chat with scientific research papers in conversational, non-technical language, enabling deep dives into research, even for individuals and groups with no science background. proem adapts to users’ language style without sacrificing insight and without resorting to misleading info that plague some AI tools’ output.

Proemial's first beta product - proem - is now released to a growing waitlist that includes people from organizations like CERN, Novo Nordisk, Google (Alphabet), Facebook (Meta), MIT, Cornell, and Amazon, whose CTO, Werner Vogels is an advisor and strong supporter of Proemial’s vision.

"One of the hardest issues Amazon engineers have to deal with is understanding which research is most applicable to the problem they are trying to solve," he says. "With their innovative use of cutting-edge AI, conversational design, and a deep understanding of science dissemination, Proemial is creating a product that will help us dig through the mountains of publications. I look forward to supporting them in bringing this to life, and I, for one, plan to become a user of their application."
 

A small group of the company’s advisors and invited users have been testing the product for some time, reporting incredible results. One advisor called co-founder and CEO of Proemial, Geet Khosla, to share her experience: "One of her family members was recently diagnosed with diabetes and prescribed a drug. When she researched this drug using Google and AI tools like ChatGPT, she got general, unreferenced answers or blue links filled with ads," he recalls. "Using proem, she got a meta-analysis conducted on 10,000 people comparing this specific drug to a few other options. For the first time in her life, she could easily understand the research papers referenced, and ask further questions to the specific paper, which helped her realize that a different drug might perform better. She sent the paper to the family doctor, who hadn’t seen this fairly new research. Her family member is now on a different, better suited drug."

The diabetes example underscores the value of what happens when people discover, understand, connect, and apply the vast deep knowledge present in scientific papers, says Khosla: "We aim to build a platform that helps us deepen our understanding by leveraging the two most powerful fundamental technologies created by humans - the Scientific Method and AI," says Proemial’s co-founder, who wants to make it easy, almost addictive, to dive deep into any topic or niche, by asking simple questions and getting trustworthy answers with references.

"We envision an exciting future, where the knowledge in research papers is actionable for everyone to apply, ushering in a future of continued progress. The solutions to some of our biggest problems are being solved in research papers that nobody reads - until now."

In December 2023, Proemial closed their pre-seed round of $2.2M, led by Danish VC, People Ventures, co-led by Dreamcraft Ventures, and supported by a group of angel investors and advisors from Apple, Google, Meta, tech leaders including Werner Vogels, CTO of Amazon, Tom Gruber, Founder/CTO of Siri (Apple), and academic luminary Serge Belongie from The Pioneer Centre for AI/ former Cornell Tech Dean.
www.proemial.ai
related items