Rivia raises €13M for agentic AI to run clinical trials

Swiss startup Rivia has raised about €13M in a new funding round to develop agentic AI for clinical trials. The Zurich-based company wants to automate not the science itself but the most painful operational layer of drug development: collecting, reconciling, checking and interpreting data that is still scattered across dozens of systems.

Why the market is broken

Clinical trials are among the most data-intensive processes in medicine. A single large Phase III study can pull in information from research sites, laboratories, patient diaries, wearables, imaging, genomics and regulatory documents at the same time. The problem is that these streams usually sit with different vendors, arrive in different formats and update at different speeds. As a result, teams spend their time not on analysis but on endless manual assembly of a single picture.

According to Rivia, the volume of data in clinical research has grown by more than 400% over the past decade, yet the underlying infrastructure has barely changed. Large systems such as Veeva and Medidata historically solved for storage and regulatory compliance, not for integrating heterogeneous sources in real time. So in practice, companies often still export files from several systems, merge them in Excel or build separate pipelines for each new protocol.

What Rivia is building

Rivia has operated since 2022. It first built a unified data layer for biotech teams and only then began adding AI agents on top of that base. In June 2024 the startup raised €3M in seed funding, and it has now closed a larger Series A round that, according to various reports, totaled $15M, or roughly €13M. Earlybird led the round; Defiant, Speedinvest, Amino Collective and Nina Capital also participated. The money will go toward growing the teams in Zurich and Boston and launching new embedded agents.

"If you don't build the data structure first, AI will work on poorly organized information and deliver unreliable results," says

Rivia CEO Erik Scalfaro.

Rivia's platform currently includes several key layers:

• a unified layer of normalized data for fragmented sources
• a library of reusable configurations for the logic of specific studies
• the Spark agent, which turns natural-language requests into clinical charts and summaries
• agents for proactive monitoring of data quality and deviations
• workflows with a transparent audit of actions that can be verified in a regulatory setting

Rivia separately points to results from a Phase 2 study in which Spark helped a sponsor's team with routine clinical review tasks, from analyzing adverse events to building cohorts and preparing descriptive summaries. According to the company, a manual task took 47 minutes on average, while a Spark response took about 2 minutes. In total that saved 91% of the time and freed up roughly 20 hours, which the team could spend on clinical decisions rather than data mechanics.

The company does not yet disclose its number of customers, but says the platform is already used in active trials.

Where agents will be useful

The most interesting part of Rivia's story is not yet another AI assistant but the attempt to embed agents in a tightly regulated environment. In clinical trials, a "smart chat" that gives a polished answer is not enough. Explainability, audit trails, validation, version control and compliance with FDA and EMA requirements all matter. Regulatory pressure is also growing: the industry increasingly needs tools that help manage risk and compliance proactively, not after the fact.

In practice, such agents can handle several expensive, slow tasks at once: flagging data quality problems in advance, finding patient recruitment risks before deadlines slip, helping teams build visualizations and summaries faster during a study, and structuring actions so they can be verified later. If this approach works at scale, the winners will be not only CROs and biotechs but drug development programs themselves, where delays quickly turn into millions of dollars in losses.

What this means

Rivia is betting on a clear thesis: the next stage of AI in healthtech is not text generation on top of chaotic files, but agentic systems that understand the logic of a clinical study and work within regulatory frameworks. The company has set an ambitious goal of cutting trial costs by up to 50%. If it can prove the reliability of this approach on real programs, the market will get not another dashboard but a new operational layer for drug development.