System is the fastest way to find, synthesize, and contextualize scientific literature, beginning in health and life sciences.
The depth and breadth of information you need to stay on top of is increasing, but the technology there to help you hasn’t changed in years. Search still means sifting through thousands of results. The more science advances, the harder the problem gets.
Plus, the way we search no longer reflects our best understanding of the complexity and interconnectedness of the systems that make up our world — from health to climate to the economy. We are driven to see silos instead of systems, impairing the quality and reliability of our decisions.
System reimagines search from the ground up, using AI to extract, relate, and visualize the findings of research. Save hours reviewing the literature — and get help connecting the dots to make better decisions.
With other search tools, you have to analyze and stitch together the literature on your own. System synthesizes the evidence for you: for a given paper, a topic or variable, or a relationship between them.
For insights to be actionable, they need to be explainable first. With System, you can see the source data for every relationship on the platform, including direct links to the underlying studies, and the statistical methods used to compute them.
Your search is usually constrained to what you already know to look for. System puts findings from studies in the context of related areas, helping you discover other factors to consider for research plans or clinical decisions.
Now it’s simple to focus your literature review on the population characteristics you care about. System starts by locating the statistical findings that match the filters in your search, so you’ll only find relevant papers.
Open new horizons for your research or clinical practice. Sign up for System today.
Unlike classic search engines or knowledge graphs which are semantically based, the atomic unit of System is the statistical relationship. Over 100 types of statistical relationships, including causal relationships, are currently recognized (e.g. odds ratio, Pearson r, and permutation score).
And with System, you get a synthesis of the statistical findings contained in and across papers, not a Large Language Model interpretation of the paper itself.
For more information on the platform, read our technical documentation.
The content on System comes from sources like PubMed. Leveraging recent advancements in Large Language Models — specifically GPT-3 — System continuously auto-extracts statistical relationships from peer-reviewed scientific papers.
To maintain quality as more information becomes available on System, users can flag any errors they identify in extraction, naming, or matching. Content that is flagged will be marked as in review and promptly reviewed by our team.
To further ensure that the information on System is trustworthy, we are continuously working to correct potential biases that may surface in our use of AI models. We publish these in our technical documentation.
System is a free, open, and living public resource that aims to explain how anything in the world is related to everything else. Our vision is to leverage real-world evidence and AI to statistically relate everything as one system. We believe that seeing the whole system — in health, climate, the economy, etc. — will help us all make better decisions: at home, at work, and as a society.
We are committed to open knowledge and open science. You can read our Public Benefit Charter here and our mission statement here.