Whitepaper

Scientific manuscript writing involves skills rarely taught explicitly in graduate programs. Existing tools treat manuscripts as generic documents rather than structured scientific arguments, creating systematic challenges in reproducibility, clarity, and data integration.

The Core Problem

Researchers face interconnected challenges stemming from a fundamental mismatch between how scientific reasoning works and how writing tools are designed.

Structural invisibility

The logic connecting claims to evidence operates implicitly. Standard word processors don't expose or support this structure. Writers must hold the entire argumentative architecture in their heads while managing formatting, references, and prose—a cognitive load that makes maintaining consistency difficult.

Data disconnection

Measurements, references, and analyses exist separately from manuscript text. Keeping them synchronized requires manual effort. When experimental values change, tracking which claims need updating becomes a manual search process, creating opportunities for inconsistencies in published work.

Collaboration complexity

Multiple authors must maintain consistency in terminology, citations, and arguments. Current tools either sacrifice real-time collaboration or require vendor lock-in—forcing a choice between isolation with control or collaboration with data traps.

Reproducibility gaps

The path from experimental data to written claim is often unclear. When values change, manuscripts don't automatically reflect updates. Reviewers questioning a measurement face difficulty tracing it to experimental context, undermining reproducibility.

These problems require different foundational assumptions about what a manuscript is and how it should work—addressing the structure of scientific reasoning itself.

Foundational Principles

Manuscripts as Structured Arguments

Scientific manuscripts connect claims to evidence through explicit reasoning. We treat this structure as a first-class concern, making it visible and manipulable rather than implicit.

Document sections represent logical components of scientific argument. Scientific entities are structured data with inherent rules. References to experimental data maintain connections, updating when sources change. This approach mirrors how scientists think about their work, not how word processors think about documents.

Data Integration Without Lock-In

Research data connects directly to manuscript text without trapping your data in our system. We store data in accessible, documented formats with comprehensive export functionality. If FirstAuthor stops serving your needs, you leave with your data intact. This commitment to portability is fundamental to how we build features.

Controlled AI Integration

AI can assist with scientific writing, but not through black-box generation that obscures reasoning or introduces errors. The role of AI in scientific writing requires careful boundaries. We use AI where it genuinely helps, but avoid applications that compromise scientific rigor.

Core Values

No User Lock-In

Your work belongs to you. We build value through useful features, not dependencies. Data lives in accessible formats with comprehensive export functionality. We compete on the value we provide.

Adding Genuine Value

We don't add features to increase engagement or create dependencies—only to address specific challenges in manuscript development. What guides development: Does this make the manuscript more correct? Does this make research more reproducible? Does this reduce cognitive load? Does this expose what was previously invisible?

Respect for Scientific Rigor

Scientific communication standards require tools that reflect those differences from creative writing. Our approach: semantic recognition respecting domain conventions (gene nomenclature, species names, statistical reporting), reference management maintaining citation context, measurement tracking preserving experimental context, and AI that suggests but doesn't assert.

We're building infrastructure for scientific reasoning, writing and publishing.