Rivia launches AI trial data review module as Scancell aims to cut cleaning to two weeks

Rivia, a Swiss clinical data company, launched DataFindings on 11 November 2026, an AI module that reviews clinical trial data continuously and automatically, and named UK cancer immunotherapy biotech Scancell as its design partner and first adopter¹. Scancell says it hopes the tool will shorten end-of-study data cleaning from eight weeks to two¹.

How the module works

DataFindings is sold as an add-on to Rivia Core, the company's platform for pulling data from separate trial systems and vendors into one structured base¹. AI agents help users draft the rules that define which data need review. After those rules are in place, the module watches the trial data on an ongoing basis and flags changes that need attention¹. Users can then review each flagged issue and hand tasks to colleagues, with every decision and action logged in an audit trail inside the platform¹.

Rivia's chief executive, Erik Scalfaro, framed the problem as one of late data rather than slow analysis: trials are often held up by data problems left unresolved, which slow the passage from trial completion to analysis and regulatory submission¹. Scalfaro said that "The increasing complexity of drug development requires a fundamentally different approach to clinical data infrastructure"¹.

Scancell has already used the module on its phase 2 data. Joe Thornton, its vice president of clinical operations, said DataFindings "has enabled Scancell to translate its phase 2 clinical trial data with greater granularity, confidence and efficiency"¹. Rivia raised $15 million in series A funding earlier this year to scale the platform and build intelligence capabilities into it¹.

What it means for the market

Rivia's case is that unresolved data issues delay the move from a finished trial to analysis and submission, so time saved on cleaning comes straight off that gap¹. Scancell's target of two weeks instead of eight shows the size of the gain the companies are claiming for a single study¹. For sponsors, the more interesting change is the model: review that runs throughout the trial rather than in a burst at the end, with rules drafted by AI agents and decisions kept in an auditable record¹. The audit trail matters for any tool touching data that will end up in a regulatory submission.

Rivia is still at the series A stage¹. A named biotech adopter gives the company a reference customer, but one partnership says little about how the module performs across larger, multi-vendor programmes run by big sponsors or contract research organisations.

Caveats

The eight-to-two-week figure is Scancell's stated aim, not a measured result¹. No data on accuracy, false flags or how often human reviewers overrule the AI-drafted rules were published with the launch, and the claims come from the company's release and a customer quote¹. Terms of the Scancell arrangement were not disclosed. Scancell's own description of its experience covers only its phase 2 data, so whether the gains hold at end of study in a larger trial is still to be shown¹.