ScienceMachine, the AI-native biotech startup building the future of automated research analysis, has raised $3.5 million in pre-seed funding. The round was led by Revent and Nucleus Capital, with support from Juniper Ventures, Opal Ventures, and select angels from the life sciences and AI ecosystems.

Co-founded by Lorenzo Sani and Benjamin Tenmann, ScienceMachine is the company behind Sam, an autonomous agent that acts as a full-stack bioinformatician. It’s built to automate scientific workflows from raw data cleaning to exploratory insight generation, embedding itself inside biotech and pharma research teams as a kind of 24/7 digital scientist.

Sam: Automating Analysis, Structuring Discovery

Sam is already being used by biotech labs to reduce data turnaround times by up to 70%, delivering structured analyses from previously siloed or unprocessed experiment data. Instead of requiring human data scientists to manually wrangle datasets, Sam interprets, cleans, visualizes, and generates key statistical summaries automatically - speeding up everything from early discovery to preclinical insight development.

The AI runs natively in scientific environments, integrating directly into experimental databases and lab notebooks. This allows it to operate like an embedded analyst, not an external interface. Sam doesn’t just summarize. It reasons. It proposes relevant metrics. It generates visualizations tailored to the underlying biology. And it adapts to feedback in real-time.

But what makes this particularly instructive for founders is how the company avoided the temptation to position itself as another AI SaaS tool. Instead, ScienceMachine treats Sam like infrastructure. They designed it to disappear - quietly doing work where the pain is chronic, not loud. It’s a strategic pattern we’re starting to see among the most high-leverage AI startups: rather than chase visibility or novelty, they go deep into underbuilt internal processes and become essential.

This approach has a second-order advantage. When a product becomes part of a team’s workflow, the switching cost increases. When it becomes part of a team’s thought process, it becomes irreplaceable. That’s a lesson every founder should remember: the most defensible products aren’t always disruptive - they’re indispensable.

Funding to Scale Team and Expand Workflow Coverage

With the new capital, ScienceMachine plans to expand its team with top-tier engineers, computational biologists, and pharma product specialists. The company is hiring with surgical precision: domain understanding is non-negotiable. Every new hire is expected to bring deep familiarity with regulated research environments and an instinct for how real scientists think and work.

The funding will also support product expansion, including support for wet-lab integration, compliance with GxP documentation standards, and broader analysis pipelines for genomics and proteomics data. By layering in these capabilities early, the company ensures Sam is future-proofed for deployment in large pharma, not just early-stage biotechs.

Traction Without Traditional GTM

What’s remarkable is that ScienceMachine accomplished all of this without a sales team and without spending on marketing. Instead, the team focused obsessively on product-market fit inside a single vertical: lean biotech teams struggling to make sense of growing data complexity.

Through inbound discovery calls and organic referrals, the company has built an early customer base that already includes multi-million-dollar labs, with a backlog of pilots lined up in both the UK and EU. Investors describe it as one of the fastest-executing technical teams they’ve ever backed.

Revent’s Rebecca Brill called it “the best example of focused, high-impact execution,” highlighting how a two-person founding team delivered enterprise-grade automation in a matter of months. Nucleus Capital added that Sam is not just another LLM wrapper, but a real example of what agentic AI in scientific environments should look like.

Building the Foundation for AI-Native Science

The broader vision is clear: as drug discovery timelines tighten and regulatory scrutiny increases, biotech companies need a better way to analyze and interpret research outputs. ScienceMachine aims to make high-quality analysis available on-demand - without needing to hire an entire data science team.

Sam isn’t meant to replace scientists. It’s designed to remove friction, eliminate repetition, and accelerate insight cycles. When labs spend less time formatting data and more time thinking about what it means, the speed of scientific discovery increases dramatically.

The team believes that the future of research won’t be “AI-assisted” - it will be AI-native. Scientists will ask questions and machines like Sam will handle the data plumbing, number crunching, and result visualization in seconds. That transformation requires trust, traceability, and technical accuracy - values ScienceMachine has embedded into its product from day one.

What’s Next

ScienceMachine is already working with its first wave of design partners to stress test Sam in live R&D environments, including oncology, rare disease, and synthetic biology. These relationships are key not just for validation, but for expanding Sam’s ability to learn scientific nuance and generate contextually relevant outputs.

The company plans to raise a larger round in 2026, after demonstrating deeper workflow automation, particularly in regulated environments. The goal is to evolve Sam from an analyst to a collaborative research teammate - capable of parsing not just data but intent.