DataJoint

DataJoint is an operating platform for scientific data, computation, and analysis that helps research labs perform, streamline, and protect their work. By combining instruments, code, data, and computation into automated workflows, DataJoint enables research that is transparent, reproducible, and AI-ready. The platform is trusted by premier research institutions including UCSF, Johns Hopkins, Harvard Medical School, the University of Houston, and Indiana University, and is backed by NIH funding totaling over $6 million in grants.

At the core of DataJoint's technology is the computational database — a fundamental infrastructure innovation that unifies every aspect of a study, including data, code, workflows, parameters, and sequences of transformations. Unlike conventional approaches, the computational database delivers flexibility without sacrificing integrity or reproducibility, making scientific processes repeatable and ready for next-generation AI. DataJoint also co-leads the SciOps discipline — alongside Johns Hopkins Applied Physics Lab and an alliance of academic and industry partners — a methodology that brings operational rigor to every stage of the research process through modular workflows, automated quality control, versioned code, and real-time collaboration around shared pipelines.

Key Platform Capabilities

• Automated research workflows: Replaces ad hoc, manual processes with structured, continuously running pipelines from data acquisition through analysis.
• Reproducibility and traceability: Records every data transformation, links data, code, and process, and enables reconstruction of any result at any point in time.
• AI-readiness: Structures data for long-term reuse and AI interpretation, ensuring inputs to AI systems are reliable, structured, and context-rich.
• Quality assurance: Eliminates errors through standardized workflows and automated QA, cutting 80–90% of time spent cleaning and processing data.
• Compliance support: Helps labs comply with NIH Data Management and Sharing Policy and supports FAIR data and FAIR workflow principles.
• Agentic AI control layer: Enables defensible and reproducible AI in regulated R&D environments, recently launched for pharmaceutical research and development.

Validated Domain Solutions

• Systems Neuroscience: Integrates physiology, imaging, microscopy, optogenetics, motion analysis, behavior, histology, morphology, multi-omics, EEG/EMG, and more into unified pipelines.
• Biomedical Moonshots: Provides stable, scalable infrastructure for science that spans decades, disciplines, and institutions, including landmark projects such as the MICrONS study published in Nature.
• Oncology: Enables research teams to unify their data, apply AI, and act faster on findings.

SciOps Maturity Model

• Level 1 – Initial: Ad hoc processes and DIY custom development.
• Level 2 – Managed: Established and repeatable processes, role specialization, and quality control.
• Level 3 – Defined: Sharable processes, open-source ecosystems, FAIR data and workflows.
• Level 4 – Scalable: Automated workflows, SciOps pipelines, collaborative environments, and teamflow.
• Level 5 – Optimizing: Closed-loop discovery with AI and human-in-the-loop systems.

SciOps Core Principles

• Modularity
• Automation
• Transparency
• Traceability
• Continuous Improvement

Research Impact and Operational Benefits

• Accelerates time to publication by months or even years.
• Enables more complex and ambitious experiments that would otherwise be infeasible.
• Allows code updates and algorithm changes without disrupting existing workflows.
• Maintains continuity as lab teams and projects evolve over time.
• Makes better use of lab time and talent by automating repetitive work.

History, Funding, and Open-Source Foundation

• Founded by Dimitri Yatsenko, a data architecture and systems engineering expert who developed the computational database concept while working in a neuroscience lab.
• Originally released as an open-source project, DataJoint Python remains freely available and gives labs a common language to describe data, code, and computational workflows.
• In 2020, NIH awarded a 5-year, $3.8 million U24 grant to disseminate the open-source framework.
• In 2022, NIH awarded a 2-year, $2.2 million SBIR grant to develop the commercial operating platform.
• The platform has been adopted by leading labs across systems neuroscience, pathology, and rehabilitation, and has enabled dozens of labs to collaborate and process petabytes of data.

DataJoint is purpose-built for scientists who need to manage complex, high-value research data without becoming IT professionals. Its open foundation ensures that anyone can read, understand, and extend a pipeline, and labs retain full ownership and portability of their data.