Reliability and privacy improvements
A consolidated production update improving responsiveness, saved-workspace consistency, analytical safeguards and Ask Grid privacy controls.
- Everyday workspace reliability. Large datasets are more responsive during ordinary block inspection and Ask Grid interactions. Expensive automatic eligibility suggestions are deferred for larger sources without preventing users from running analyses.
- Everyday workspace reliability. Workspace switching, saved layout and zoom controls, and result-freshness verification are more consistent across reopening and reloads.
- More dependable analysis actions. Predictor eligibility previews and nested-filter counts now better agree with the data actually used for an analysis.
- More dependable analysis actions. Stored-filter reruns include stricter row-alignment safeguards to avoid silently pairing the wrong observations.
- Ask Grid privacy choices. Manage whether questions and responses are shared with Grid for product improvement, and review previously shared items. Sharing is controlled by the user.
Ask Grid: contextual questions
Completed the initial evidence-grounded Ask Grid experience. It became available in production with the November update.
- Questions in the context of analysis. Ask about findings, statistical measures, method selection, diagnostics and what a result means using information from the current analysis.
- Questions in the context of analysis. Relevant question prompts and follow-ups help users investigate results without leaving the workspace.
- Answers grounded in evidence. Deterministic answers use recorded analytical evidence and respect missing, outdated or unavailable results; they are not presented as open-ended generative AI.
Portable workspaces (.gridw)
The first portable .gridw workspace export was
completed, establishing a way to preserve and move an investigation.
- Workspace archives. Export and import analytical workspaces using
.gridwarchives, including the records needed to reconstruct the investigation. - Workspace archives. Import checks and source-resolution workflows help identify datasets that must be matched or supplied before analyses can be rerun.
Portable datasets (.gridd)
Completed the first .gridd dataset download, adding
a portable way to retain structured data.
- Dataset portability. Download datasets as CSV or Grid’s portable
.griddformat. - Dataset portability. Import
.gridddatasets and download saved datasets without needing to open them for editing first.
Statistical reliability improvements
Strengthened analytical behavior across difficult inputs and clarified when results need caution.
- Statistical edge cases. Improved handling of nulls, filtered populations, paired observations, prediction inputs and unsupported cases.
- Statistical edge cases. Added checks for filtered rows, paired observations and consistent analytical preparation.
- Clearer result limitations. More consistent explanations distinguish usable findings from results that cannot be safely calculated or interpreted.
Clearer findings and next steps
Improved the way Grid explains statistical findings and suggests relevant next steps.
- Understand the result. More structured explanations of method selection, diagnostics and key evidence across Describe, Relate, Compare, Associate and Predict.
- Understand the result. Capability-aware suggestions help distinguish suitable follow-up analyses from unavailable or unsupported actions.
- Navigate the investigation. Minimap and viewport refinements make larger analytical workspaces easier to understand and navigate.
Production backups established
The first production backup workflows ran successfully on September 10; the first successful scheduled run followed on September 11.
- Database and dataset protection. Production backup workflows cover PostgreSQL records and referenced R2 dataset objects, with downloadable manifests and integrity checks.
- Database and dataset protection. Recovery tooling and retention planning were established to support continued protection of saved datasets and workspaces.
- Operational continuity. The successful manual runs were followed by the first successful scheduled production backup on September 11.
Dataset and workspace refinements
Refinements to importing, organizing and revisiting data during early private-beta development.
- Handling and organizing data. Stronger processing for large imported datasets and more resilient descriptive summaries.
- Handling and organizing data. Clearer project usage, trash controls and dataset-dependent freshness tracking.
- Getting started. Refined introductory guidance, sample workflows and workspace interactions for new users.
Telene Grid opens
The production website opened for invitation-based private beta. The first invited users joined later.
- Bring data into Grid. Import CSV and Excel files, connect supported cloud files, start with blank or demo datasets, and edit or derive columns in the table.
- Investigate visually. Explore descriptive statistics, relationships, group differences, associations and prediction models through connected analytical blocks.
- Investigate visually. Inspect supporting details and diagnostics, build branches with filters and groups, and arrange analyses using Tree, Rows or Lanes layouts.
- Keep and communicate your work. Organize datasets and saved workspaces within projects, prepare report views, and share supported workspace views with configurable access.