AIPOCH announces the public release of AIPOCH Open-Science, an open-source AI research workbench designed to streamline planning, execution, and sharing of AI-assisted research through portable .science packages and a unified workspace.

The project involves a broad international collaboration across 76 institutions, including Zhongshan Hospital of Fudan University, Amsterdam UMC, Beth Israel Deaconess Medical Center, Peking University, Zhejiang University, UIUC, Technion, SenseTime, and Amoy Diagnostics, supporting validation of the Open-Science approach.

Open-Science aims to address fragmented workflows, context transfer, model/vendor lock-in, and opaque outputs by providing a persistent, inspectable research environment that unifies literature review, data analysis, visualization, reporting, and execution within a single workspace.

The platform includes 23 built-in Skills, 525 marketplace Skills, 36 research connectors, and four agent frameworks—Claude Code, OpenCode, Codex, and CodeBuddy—supporting multiple model providers and execution environments.

The Skill Marketplace features 525 Skills and 10 Specialists, enabling a global open-source community to contribute and reuse research capabilities within the workbench.

AIPOCH hosts a Skill Marketplace with 525 Skills and a Specialist Marketplace with 10 Specialists, fostering community contributions and reviews to expand reusable research capabilities.

Key capabilities include portable .science packages, an integrated literature library with citation support, open-source and model-agnostic architecture, and locally stored project data with configurable access and external model integrations.

Key features include portable .science packages, an integrated literature library with citation tools, open-source and model-agnostic architecture, and locally stored project data with configurable access and external model integrations.

The platform targets diverse scientific and technical fields, including life sciences, bioinformatics, statistics, machine learning, chemistry, physics, and environmental science, and is intended for researchers, students, educators, developers, engineers, and analysts.

Designed for broad scientific and technical workflows across life sciences, bioinformatics, statistics, ML, chemistry, materials science, physics, and environmental science, the platform targets researchers, students, educators, developers, engineers, and analysts.

The platform supports multiple AI model providers and configurable execution environments, with integrated code execution (Python, R, shell) and traceable artifacts linking outputs to inputs, code, and execution history.

Further information and source code are available at aipoch.com and GitHub (github.com/aipoch/open-science).