Introducing the Satya Research Platform: Making Psilocybin Research Easier to Explore

For the past several years, Satya Therapeutics has been known primarily for one thing: providing legal, regulated psilocybin services in Oregon.

But operating a psilocybin service center has continually brought us back to a more fundamental problem.

The number one thing we have learned from people calling us interested in doing a journey is that the average client has done a tremendous amount of research: reviewed clinical research, read books like How to Change Your Mind, studied service center websites.

The science is difficult to navigate.

There is a rapidly growing body of research involving psilocybin and psilocin. There are published studies, reviews, clinical trials, trial registrations, different therapeutic applications, different populations, different doses, and sometimes conflicting conclusions.

For someone simply trying to understand what the research actually says, it can be overwhelming.

So we decided to build something to make it a little easier and more comprehensive than Google searches or AI searches.

Introducing the Satya Research Platform

We have launched the first version of the Satya Research Platform, an experimental research tool designed to make the scientific literature surrounding psilocybin and psilocin easier to explore.

Try the research platform:

research.satyatherapeutics.com

Instead of beginning with a conventional search box and trying to determine the right keywords, users can simply ask a question.

Questions such as:

What does the research say about psilocybin and depression?

What has been studied regarding psilocybin and end-of-life anxiety?

What clinical trials involving psilocybin are currently registered?

The platform searches Satya’s curated research collections and provides an answer, including source documents (the publications themselves).

There is an important difference between this and simply asking a general-purpose AI chatbot a question about psilocybin.

We wanted the evidence to remain visible.

Building From a Defined Body of Evidence

The current Satya system uses two distinct scientific collections.

The first contains approximately 3,500 curated publication records from PubMed, the biomedical literature database maintained by the U.S. National Library of Medicine.

The second contains relevant registered studies from ClinicalTrials.gov.

We deliberately keep these collections separate.

A published scientific paper and a registered clinical trial aren’t the same thing.

A trial can be registered before results have ever been published. The trial may even have been cancelled. A publication can discuss findings from a completed trial. A registered study can be underway for years before a paper appears—or may never result in a publication at all. We think being able to see what is on the horizon is as important as what has been studied already.

The Satya platform treats published literature and registered clinical trials as separate evidence streams rather than throwing everything into one enormous pile of documents.

The underlying records point users back to their original public sources.

AI Is the Interface, Not the Evidence

Large language models can produce remarkably convincing answers. But a convincing answer isn’t necessarily a reliable answer. AI models make a host of errors:

• Misses relevant studies — Important studies can be in the database but fail to appear in an answer.
• Incomplete source data — Some records have no abstract, so the AI may have too little information to recognize or evaluate an important study.
• Misses important findings — The AI may retrieve the correct study but fail to include significant outcomes reported in it.
• Doesn’t always weigh evidence appropriately — Case reports, reviews, clinical trials, and other evidence types can receive inappropriate prominence.
• Citation problems — Relevant research may be identified without displaying the correct PMID or other source information.
• Limited by abstract-level evidence — An abstract may omit important methods, outcomes, adverse events, follow-up data, or limitations contained in the full paper.
• Presentation errors — Formatting or character-encoding problems can occasionally distort things such as units or symbols.
• Can sound more certain than it should — A fluent, convincing answer can still be incomplete or wrong.
• The corpus itself is bounded — If Satya hasn’t acquired a study, the AI cannot use it; absence from an answer does not mean the research doesn’t exist.
• AI answers require verification — The platform should be treated as a research-navigation tool, not as medical advice or an authoritative final interpretation of the evidence.

We’ve Improved the Approach

We aren’t asking AI to simply tell us what it knows about psilocybin.

We built a bounded research environment in which AI helps people navigate a scientific collection that Satya has deliberately acquired and organized.

That doesn’t eliminate uncertainty.

It doesn’t make every answer correct.

And it certainly doesn’t turn an AI system into a physician, researcher, or substitute for professional medical advice.

What it does is make a complicated body of scientific information substantially easier to interrogate.

That is already very helpful, as our facilitators have access to the research as well, making them better able to direct clients to possible answers to their questions.

Why Satya Is Building This

Building this platform has also forced us to think more deeply about what Satya itself is becoming.

We started with psilocybin services.

Our service center represents one way of translating emerging scientific knowledge into something people can actually use: a structured environment in which adults can legally access psilocybin services within Oregon’s regulated system.

The research platform represents another.

And that has led us toward a broader statement of what we’re trying to accomplish:

Satya takes complex scientific evidence and turns it into evidence-based tools that help people explore better health outcomes.

This Is a Beta

The research platform is still early.

We want to discover what people actually ask.

We want to find places where the underlying research collection is incomplete.

We want to see where answers are useful—and where they’re not.

We want to improve how studies are presented and cited.

And, perhaps most importantly, we want to understand whether making scientific evidence easier to interrogate actually helps people understand complicated health questions.

The platform itself will evolve as we learn.

The underlying research collection will evolve too.

One particularly important limitation is that the current publication database primarily contains bibliographic information and abstracts supplied through PubMed. It should not be mistaken for a comprehensive full-text archive of the scientific literature.

That’s one of many things we’re thinking carefully about as the platform develops.

An Experiment in Making Science More Usable

There is an enormous amount of scientific information available today.

Paradoxically, access to more information doesn’t necessarily make it easier to understand what we know.

What evidence exists?

How strong is it?

Where did it come from?

What remains uncertain?

And what should we conclude from it?

The Satya Research Platform is our first attempt at building a better way to explore them.

For now, we’re beginning with the field we know best: psilocybin.

And we’re inviting people to use it, challenge it, ask difficult questions, and help us discover what it should become next.

Try it here:

research.satyatherapeutics.com