About
Built by people who watched MSLs spend their days searching PDFs
Argon AI was founded in Boston in 2024. A small team with backgrounds in clinical informatics, natural language processing, and pharmaceutical operations, working on one problem: medical science liaisons should not have to spend their expertise on document retrieval.
Founding story
The problem was not knowledge. It was retrieval.
Samy Danesh spent several years working in medical affairs technology at a pharmaceutical software firm before founding Argon AI in 2024. In that role, he worked alongside medical science liaisons who held doctorate-level expertise in pharmacology, oncology, and rare disease. What he watched was not people struggling to understand the science. It was people spending 30 to 40 percent of their working day searching for documents they knew existed.
An MSL preparing for a KOL visit would spend three hours searching SharePoint, emailing a colleague in medical information, and manually cross-checking a 90-page prescribing information PDF to confirm a single dosing exception. The knowledge was there. The retrieval infrastructure was not.
Argon was built to close that gap. Not as an autonomous AI that generates answers without sources, but as a retrieval system that returns the specific paragraph the answer comes from and puts it in front of the person who needs to verify it before using it in a scientific exchange.
The company is angel-backed and based in Boston's Seaport district, close to the concentration of pharmaceutical operations and clinical informatics talent in the Massachusetts ecosystem.
Team
The people building Argon
Backgrounds in clinical informatics, natural language processing, and pharmaceutical operations. Small by design while we are still learning what the best medical-affairs teams need from a retrieval tool.
Samy Danesh
CEO and Co-Founder
Previously in medical affairs technology at a pharmaceutical software firm. Saw the retrieval problem up close and built Argon to address it. Based in Boston.
Priya Nair
Head of NLP
Specializes in retrieval-augmented generation and biomedical document understanding. Her work on clinical text parsing is the core of Argon's source-grounding capability.
Marcus Reyes
Head of Product
Background in pharmaceutical operations and clinical data management. Brings the MSL workflow perspective that keeps Argon's product decisions grounded in how medical-affairs teams actually work.
How we think about building AI for a regulated industry
A few principles that have not changed since day one.
The source is not optional
An AI answer without a source citation is not an Argon answer. The source is part of the answer, not a bonus feature. We built the citation into the data model, not the display layer.
A human reads the source before it reaches an HCP
Argon is decision support, not a decision-maker. The MSL reads the source. The medical officer reviews the response. The physician makes the clinical decision. We build for that flow, not against it.
Say what you do, not what sounds good
We do not display HIPAA or SOC 2 badges we have not earned. We describe exactly what the architecture does. When we have a third-party audit, we will share it. That is the only credibility signal that matters in this industry.
Want to talk to us directly?
We are a small team and we respond to every serious inquiry. If you want to ask how the retrieval architecture handles clinical study reports specifically, or what our BAA process looks like, we will answer directly.