01
No heavy coding required
Choose variables, conditions, and statistics in the UI. Nothing is hidden: the generated R program is written next to every output you create.
Why it exists
Statisticians should spend time on the analysis, not the plumbing.
Clinical TLGs are routine in structure but expensive to program. CLAVIS turns the configuration of a display into a visual task: pick the population, treatment columns, and statistics you want, and the app prepares the data, computes the display, and produces an RTF output, a QC transport dataset, the R program that ran, and its execution log.
UIVisual builders for tables, listings, and graphs
ComputeAll statistics in open-source R
OutputRTF + QC XPT + R program + log
Templates73 table shells, 50 listing shells, 16 graphs
Core capabilities
Visual configuration, reproducible open-source computation.
02
Visible development steps
Data preparation, derivations, subsetting, and each display module are separate, inspectable steps you can preview before committing to an output.
03
Re-usable templates
Save any configuration as a template or spec and reuse it across outputs, cycles, and studies. A standard shell library ships with the app.
04
Metadata-driven titles
Load a planning workbook once and every output takes its output number, titles, and footnotes from it, with no drift between plan and deliverable.
05
Submission-style deliverables
Paginated landscape RTF with titles, footnotes, page numbering, and an audit footer, plus a QC XPT dataset carrying underlying cell values.
06
Open-source statistics
Computation runs on R and established CRAN packages, so results are reproducible by anyone with R and the generated program.
Built-in modules
Four builders around one reproducible pipeline.
TBL
Table Builder
Summary tables: descriptive statistics, categorical and numeric frequencies, worst-grade hierarchies, shift tables, survival estimates, and comparative tests.
LST
Listing Builder
Subject-level listings with multi-line composite columns, page grouping, sorting, and controlled pagination.
GRP
Graph Builder
Kaplan-Meier curves with risk tables, subgroup forest plots, and waterfall plots of subject-level change.
EXP
Data Explorer
Inspect datasets, variables, and distributions, and prototype derivations before building an output.
Try the demos
Two complete demo workspaces ship with the app.
Open the Hub in the deployed app and load a workspace profile — no data of your own is required.
01
CDISCPILOT01 Demo
Eight safety tables and four listings over the public CDISC SDTM/ADaM Pilot Project data.
02
DEMO001 Demo
Kaplan-Meier, forest, and waterfall figures over a synthetic 100-subject, two-arm dataset.
03
Inspect first
Pre-generated RTF, XPT, R programs, and logs ship under each demo output folder.
04
Reproduce
Press Compute then Generate Output in a builder to recreate the deliverables yourself.
Development status
Active development, provided for testing and demonstration.
CLAVIS is under active development and is not validated, qualified, or operated in a regulated environment. It is provided for testing, learning, and demonstration at your own discretion, and is not a substitute for validated programming of regulated deliverables.
CLAVIS 1.1.0 is developed by Snowbird Analytics. Built-in demos, template libraries, and the full R computation layer let you evaluate the workflow before committing to a deployment.
- Python service (standard library only) plus an R computation layer.
- Standard libraries: 73 table shells, 50 listing shells, 16 graph templates.
- Demo specifications: 8 tables, 4 listings, 3 figures.
- Reusable across outputs, cycles, and studies via saved specs.
Want to evaluate CLAVIS for your team's TLG workflow?
Try the live app with its bundled CDISC and synthetic demos, or talk to Snowbird about deploying CLAVIS in your own environment.
Open the CLAVIS App