Research software and training, built by researchersAuthentic Research Partners · Princeton, NJ
Research software, data workflows, and AI capability building for pharma, research institutions, and R&D organizations. Built to last and run on your hardware and infrastructure.
What we solve
We've spent a decade at the intersection of research and software engineering.
We understand the science, we build production software, and we start by understanding the actual problem before writing any code.
"It works on my laptop"
Your scientist wrote a Python script that does exactly what's needed. Now three people need to use it, IT can't support it, and the author is moving on.
The spec was right. The tool was wrong.
Engineering delivered what was asked for. Scientists tried it twice and went back to spreadsheets. The problem wasn't the build. It was what got lost in translation.
Everyone says AI. Nobody says which model, where, or why.
Your team is evaluating tools, models, and frameworks without a clear picture of what fits your actual research workflow. The wrong choice wastes months. The right one changes everything. The difference requires understanding both the science and the technology.
The vendor built a demo, not a product.
It looked great in the presentation. In production, it doesn't handle your data, your edge cases, or your compliance requirements.
The postdoc who built everything is leaving.
Undocumented code. No tests. No one else knows how it works. Six months of work is about to walk out the door.
Research and AI Capability Building
We train both research and technical teams. For technical teams, that means understanding research workflows, scientific reasoning, and how to build better tools for researchers. For research teams, that means learning how to use AI in meaningful ways to support scientific work. This grows naturally out of the same problem we solve in product development: connecting research needs with the right technical execution.
Tools & Research Initiatives
research-mentor
scicode-lint
sciwrite-lint
Society of Teen Scientists
AI Meets Science Conference
AI for Scientific Research
Managing Team
Research scientists, enterprise engineers, and education leaders. Combined: 60+ years in R&D, 21 patents, systems serving Fortune 500 to national labs, and programs that have trained hundreds of students and professionals.
Sergey Samsonau, Ph.D.
Physics PhD. 10+ years building AI and research software at the intersection of science and technology. Led technical design of NYU's university-wide GenAI ecosystem. Created AI capability for a radiology company. ML solutions recognized for innovation at Credit Suisse. Multiple peer-reviewed publications. Former faculty at NYU Tandon and CUNY.
Olga Vine
20+ years across product management, engineering, and technical leadership. Enterprise systems delivering multi-million-dollar projects in regulated industries (healthcare, banking, government). Hyperscience, CIVIE, Vroom (scaled engineering through IPO).
Dr. Luke Perkins
30 years in R&D program management and technology commercialization. Schlumberger and Princeton Plasma Physics Laboratory. 21 patents, 100+ invention disclosures. Led PPPL's $27M FLARE fusion facility. Now advises on R&D strategy, technology commercialization, and research program design.
Matthew Pearce
Matthew has over 30 years of experience advancing research-based education at some of the world's leading high schools: Latymer Upper School, Thomas Jefferson High School for Science and Technology, and the Princeton International School of Mathematics and Science. A pioneer in bringing authentic research into secondary education, he has mentored countless students who went on to excel in science, engineering, and innovation.
Open a Dialogue: hello@arpconnect.com