Everything between your dataset and your conference poster

One upload becomes verified results, a shareable dashboard and a print-ready poster. Every number is computed by code; AI only writes the words, and a verifier checks each number in the text against the computed values.

Data in

  • Upload CSV, Excel (all sheets checked), SPSS .sav with value labels, or Stata .dta.
  • Import directly from REDCap and KoboToolbox; refresh daily, weekly or monthly.
  • Automatic profiling: column types, missing values, Likert scales, time ranges, and a cleaning report of everything trimmed or dropped.
  • Privacy check for names, e-mails, phone and national ID numbers; identifier columns are never analysed or published.

Analyses

  • Descriptives and Table 1, Welch t-test, Mann–Whitney, Welch ANOVA with Games–Howell, Kruskal–Wallis with Dunn (Holm), paired tests, chi-square and Fisher's exact test, Pearson and Spearman correlation, linear and logistic regression, Cronbach's alpha, Likert distributions, trends, XmR control charts and bed-capacity KPIs.
  • Assumption checks with every result (normality, equal variances, expected counts, multicollinearity, sample size).
  • The exact Python code for each result, downloadable and re-runnable.

Written results you can trust

AI drafts the Results text with placeholders instead of numbers. The app fills in the computed values in APA 7 style and an independent verifier checks every number, every direction word (higher/lower) and every significance claim. Each section is marked Verified or Needs review, in English or Arabic.

Share

  • Interactive dashboard as one self-contained file: works offline, aggregates only, small counts suppressed, optional password.
  • A0/A1 posters in portrait or landscape, English or Arabic, with a QR code that opens the live dashboard and counts scans.
  • REST API, webhooks and a remote MCP server so assistants such as Claude, ChatGPT and Gemini can run your analyses.

Frequently asked questions

Does the AI calculate any statistics?

No. Every statistic is computed by published Python libraries (pandas, SciPy, statsmodels). The AI only writes sentences around placeholders, and the verifier rejects any number that does not match.

Can I see the code behind a result?

Yes. Each result card shows the exact code that produced it, with library versions, and you can download it as a script.

Is my data sent to the AI?

No rows are ever sent. The AI receives column names, summary statistics and the names of computed values.

Try it with your data See a sample dashboard