Product tours
Six video series — 33 chapters, about 45 minutes end to end — showing a systematic review run from a plain-language question to a graded result, and the assistant that answers questions about the evidence once it is in. Every series is filmed on the same worked example you can open yourself, read-only, no account.
The whole pipeline, in three and a half minutes.
One continuous pass over a published meta-analysis reproduced end to end — search, screening, extraction, appraisal, pooling — before the series below slow each stage down.
Configuring a review
From a plain-language question to a locked protocol: PICO and eligibility, the search strategy, screening policy, the bias instrument, endpoints, and sign-off.
Search and screening
The locked search run against PubMed and ClinicalTrials.gov, every record screened with its reasons on record, the full texts retrieved, and each one screened again in full.
Data extraction
Every number pulled from the papers carries its quote, its place in the paper, and its history — and what is missing is counted, not hidden.
Risk of bias
RoB 2.0 judgements as a traffic-light matrix, each cell opening to its written reasoning, with the instrument choice itself a recorded decision.
Statistical analysis
Ask in plain English for pooling, heterogeneity, publication bias or a chart — every run kept on record and repeatable.
Ask about your evidence
A read-only assistant that answers from the review’s own data — coverage, gaps and the rows behind every figure — then the same questions asked across a whole portfolio of reviews, with the answers it refuses to give.
Seen enough to want the controls?
The worked example behind every one of these tours is open read-only — the search, the screening decisions, the extracted values and the pooled result, every number traceable to its source.