CALL FOR TOOLS

PRIORITY INITIATIVES / 2025

We're seeking innovative tools that transform how science is done. These four initiatives represent opportunities to fundamentally improve scientific productivity and reproducibility.

Make Science Declarative

Much of the scientific software workflow is still imperative, requiring scientists to repeatedly write complex and error-prone algorithms. Over the past decade, software engineering has moved to declarative tools—Kubernetes, Docker, React, Terraform—and reaped huge benefits. Science should do the same. Declarative science would make experiments more reproducible and researchers more productive.

Example Areas

Parameter-driven object instantiation, declarative experiment orchestration, scientific DSLs, constraint-based specification

Make Science Intelligent

Current AI tools often fail in scientific contexts because they ignore the unique requirements of research— rigor, reproducibility, and domain constraints. We need thoughtfully designed interfaces where AI enhances scientific reasoning rather than replacing it, building trust through uncertainty quantification and interpretability.

Example Areas

AI-assisted hypothesis generation, intelligent literature synthesis, automated method tracking, codebase generation from papers

Make Science Measurable

The vast majority of scientific process—failed experiments, negative results, dead-ends, and tacit knowledge—vanishes without a trace. We need open protocols for scientific quantities, inspired by OpenTelemetry's success in software. Such protocols would capture this "dark data" and enable interoperable measurement across instruments and experiments, turning informal knowledge into searchable, reusable assets.

Example Areas

Open telemetry protocols for science, negative result databases, automatic experiment logging, dark data capture systems

Make Science Asynchronous

Science remains stubbornly synchronous and location-bound. Critical discussions happen in hallways, brilliant minds outside major hubs are excluded, and collaboration requires physical presence. We need GitHub-like infrastructure and async communication tools for scientific knowledge—branching, merging, version control, and persistent discussions that transcend time zones.

Example Areas

Scientific communication platforms, protocol versioning systems, distributed peer review, async experiment coordination, persistent lab notebooks

Submit a Proposal

We're actively seeking proposals from researchers, engineers, and teams who understand the unique challenges of scientific computing. We provide mentorship, resources, and potential funding for promising projects.

Submit your proposal