Soulmates Ventures Backs FluoSphera Rethinking Preclinical Drug Development

• With Soulmates Ventures’ support, Fluosphera is building scalable, human-relevant preclinical models to reduce drug development failure
• Its multiplexed platform enables early, decision-grade insights across multiple interacting human tissues
• The company balances scientific rigor, scalability, and real-world pharma adoption through evidence-led integration
• The newly raised EUR 1.23M support commercial growth, automation, and broader acceptance of non-animal testing

This January, the well-known Czech VC firm Soulmates Ventures (invested in Rightcharge, among others) led the EUR 1.23M round of FluoSphera, the Swiss biotech startup developing its next-generation of human-based preclinical trials platform. Other investors included IndieBio New York and an unnamed angel investor from Switzerland.

Founders and Their Shared Realization of Preclinical Models Failing Humans

FluoSphera was founded in 2021 by Dr Clélia Bourgoint (CEO), Aurélien Roux (scientific advisor), and Dr Gregory Segala (CSO). The team brings together complementary expertise across biology, biotechnology, and biophysics. Dr Segala comes from a profound background in oncology, Mr Roux brings a strong academic and biophysical perspective, and Dr Bourgoint’s background is in biotechnology and advanced microscopy, with experience in drug screening and translational research.

From different angles, the co-founders saw how poorly current preclinical models could predict what happens in humans. After seven years researching new breast cancer treatments, Dr Segala led a synthetic biology team at the iGEM competition, where the team developed the idea of combining multiple human tissues in vitro. The goal was to mimic human biology more closely and better predict how drugs behave before entering clinical trials.

The Brutal Math of Drug Development, the Shortcomings of Today’s Preclinical Tools, and the Imperfection of Animal Models

To reiterate, developing a new drug today is a long, expensive, and remarkably uncertain journey. On average, it takes 10 to 15 years and requires investments of USD 2–3B to bring a single therapy to market. Despite this enormous effort, more than 90% of drug candidates still fail during clinical trials. In oncology, the odds are even worse, with success rates falling below 5%. These failures reveal a systemic problem in how drugs are evaluated long before they ever reach patients.

At the heart of this issue lies the limited predictive power of today’s preclinical models. Traditional 2D cell cultures are widely used but oversimplify human biology, failing to capture interactions between organs and tissues. More advanced 3D models offer greater biological relevance, yet their complexity, high costs, and demanding maintenance put them out of reach for many research teams.

Animal models, on the other hand, while deeply embedded in drug development, also fall short. They cannot fully replicate human biological responses and raise ethical concerns, all while contributing to late-stage failures. The consequence is wasted time, inflated R&D costs, and delayed access to potentially life-saving treatments. The growing gap between scientific effort and clinical success underscores an urgent need for scalable, affordable, human-relevant approaches to preclinical testing.

From Academic Proof to Industrial Reality

FluoSphera started as a scientific idea in academia, where the team first proved that combining multiple human tissues could change how drugs behave compared to single-tissue models. Early on, it became clear that this approach could unlock much more predictive insights.

‘A major milestone was stepping out of academia to build something for real-world use. Another turning point was showing that the platform could be simple and scalable, without relying on complex chip-based hardware. That was critical to make it practical for pharma companies and CROs. Also, working with our first commercial partners and seeing them come back for repeat studies has been one of the strongest signals that we’re solving a real and urgent problem,’ Dr Bourgoint recalls.

Today, FluoSphera offers a multiplexed in vitro platform designed to assess drug behavior in a way that closely reflects human biology, helping developers identify the most promising candidates much earlier in the discovery process. Rather than testing tissues one by one or relying on complex and costly organ-on-chip systems, the company’s patented technology brings up to six or seven different human tissue models together within a single well. Each tissue is labeled with proprietary fluorescent markers, allowing researchers to monitor them independently while observing how they interact as part of a connected system.

The combination of human relevance and relative affordability comprises FluoSphera’s competitive edge. By evaluating multiple tissues simultaneously, the platform accurately captures organ-level responses and systemic effects that are often missed in isolated models. It also significantly improves efficiency: researchers can assess both therapeutic efficacy and potential side effects in a single experiment. Crucially, this gets done early in development, reducing late-stage failures and helping avoid unnecessary animal testing further down the pipeline.

Practical Precision and Actionable Tissue Diversity

One of the issues that arise, nevertheless, is to incorporate tissue-level diversity—sex, age, disease state, metabolism—realistically into scalable drug testing. This testing needs to stay standardized, reproducible, and decision-useful. For the FluoSphera team, the most realistic route is a panel-based approach that mirrors how clinical development already works. In this approach, a candidate gets tested across a curated panel of human-relevant tissues that represent meaningful sources of variability such as sex, age, disease state, and metabolic capacity.

In practice, this means combining donor- or patient-derived biology, including iPSC-derived tissues when appropriate, with metabolism-competent components so that exposure, bioactivation, and detoxification are captured instead of being assumed. It also means controlling the microenvironment and using co-culture formats when they are needed to reproduce tissue behavior that is otherwise lost in simplified systems. Importantly, these models must remain compatible with automation and high-throughput workflows, and they must generate quantitative, interpretable readouts rather than qualitative observations.

‘Scalability ultimately comes from rigor and modularity: standardized protocols, strict quality control, assay qualification, reference compounds, and data structures that allow performance to be compared across tissue variants and cohorts. Over time, this creates a dataset that supports stratified prediction, linking response patterns to underlying biology and metabolism. That is how tissue diversity becomes actionable for precision medicine: not as unlimited complexity, but as validated, representative diversity that helps developers identify who is likely to benefit, who may be at risk, and how to adjust indication or dosing earlier and more confidently,’ Dr Bourgoint explains.

Scaling Without Overclaiming

A tension between scientific rigor and commercial speed is not uncommon. FluoSphera team’s rule of thumb is that a platform is ‘ready enough to scale’ when it reliably fits into a real customer decision better than existing options, and does so with a level of robustness that holds across operators, days, and relevant sample variability.

‘For us, readiness isn’t defined by perfection. It’s defined by whether the output is trusted enough to influence an actual development choice, such as a go/no-go decision, lead prioritization, or risk mitigation strategy in a partner program. Real-world use cases are therefore central: if a partner can run a study, interpret the results, and change what they do next because of the data, that is a strong signal we are ready to scale,’ Dr Bourgoint clarifies.

In parallel, the team runs continuous validation in a structured way. Assay performance gets qualified with clear acceptance criteria that support screening-quality statistics. The platform’s boundaries of applicability get defined early and refined as datasets grow, so scaling does not mean over-claiming. In other words, the FluoSphera team knows it’s time to scale once the platform is decision-grade for a defined set of use cases, and they keep deepening validation as they expand to new indications, modalities, or more complex biological contexts.

Overcoming Pharma Inertia with Side-By-Side Proof

Breakthrough tools often struggle with adoption. In pharma, from Dr Bourgoint and her team’s experience, the resistance is usually not ideological—it’s operational. Animal workflows are embedded in established decision gates, internal benchmarks, and team ‘muscle memory,’ so a new approach often feels risky: people worry about comparability to historical data, how to interpret the readouts, and whether switching could slow programs or create uncertainty in governance.

To tackle this, FluoSphera avoids a ‘replace everything’ message. In line with the FDA’s direction toward broader acceptance of New Approach Methodologies, the company starts with side-by-side studies against the current standard using reference compounds and the partner’s own historical benchmarks, and the parties agree upfront on what success looks like in decision terms. When it becomes apparent to teams that the data is reproducible, maps to their existing gates, and strengthens go/no-go confidence, adoption follows naturally.

‘For market education, we keep it evidence-led and practical: we share validated use cases, define where the platform applies and where it doesn’t, and translate performance into what matters to each stakeholder—better early decisions, fewer surprises later, and faster, more confident prioritization,’ Dr Bourgoint adds.

FluoSphera is already working with biotechnology and pharmaceutical companies, with particularly strong adoption in antibody–drug conjugate (ADC) development. The company also partners with Revvity, a global leader in high-throughput screening, to support scalable drug testing workflows. By improving early decision-making and reducing late-stage failures, FluoSphera’s platform can help developers save an estimated USD 100–500M per drug candidate. In parallel, the technology supports compliance with evolving regulatory frameworks, including the FDA’s Modernization Act 3.0, which promotes validated non-animal testing approaches in preclinical research.

The Funding and What It Unlocks

For Soulmates Ventures, FluoSphera stands out for its exceptional founders and groundbreaking science.

‘As the pharmaceutical industry transitions away from animal testing, the need for reliable human-relevant models is immense. FluoSphera is opening a multi-billion-dollar opportunity by giving drug developers the means to innovate faster, safer, and more ethically. It represents a completely new approach to developing the next generation of medicines,’ the fund’s founder and managing partner Hynek Sochor states.

The funding will primarily support FluoSphera’s commercial expansion and platform scalability. This includes strengthening business development efforts with pharma companies and CROs, as well as expanding automation and AI-driven image analysis. In two to three years, the team expects its success to reflect in a growing base of pharma and CRO partners, studies run at much larger scale, and, importantly, clear evidence that FluoSphera’s data is used to inform real go-or-no-go decisions in drug development.

The Future of Drugs and Beyond

In conclusion, Dr Bourgoint envisions a future where regulators accepted advanced in vitro systems as primary decision-making tools:

‘It would fundamentally change when and how decisions are made. Instead of advancing compounds far into development before discovering systemic issues, developers could identify risks and opportunities at the earliest stages. This would reduce late-stage attrition, shorten development timelines, and allow resources to be redirected toward the most promising programs. In practice, this shift could help drug developers save USD 100–500M per molecule by avoiding costly late-stage failures. It would also accelerate innovation by making it safer to explore new modalities and mechanisms. Ultimately, it would mean fewer animals, better science, and faster access to effective therapies for patients.’

Beyond drug discovery, FluoSphera detects its technology’s potential to contribute to a much broader understanding of human biology. Similar to how Illumina made DNA sequencing affordable and widely accessible, FluoSphera is armed to do the same for advanced human tissue modeling. Over time, this could also enable more nuanced, semi-personalized approaches to medicine, where drugs are tested across representative population subgroups defined by factors such as age, metabolism, or comorbidities, bringing the field closer to true precision medicine.

Addressing one of the most persistent bottlenecks in drug development, Fluosphera replaces poorly predictive models with scalable, human-relevant biology that can meaningfully guide early decisions. By making systemic human testing practical, reproducible, and compatible with real-world pharma workflows, the company is helping shift drug discovery from late-stage failure toward earlier confidence. With wider adoption, this approach has the potential to deliver better medicines faster, reduce reliance on animal testing, and fundamentally raise the scientific standard of how new therapies reach patients.