Member Technical Staff - AI​/ML Scientist

Member of the Technical Staff - AI/ML Scientist

We’re leveraging AI to solve one of the most consequential challenges in the pursuit of technical progress: the distribution of scientific innovations to the real world. We’re well‑funded with top investors and are building a world‑class team.

For innovations to have broad impact, they must scale beyond the lab. For too long, the transfer of discoveries from lab bench to real‑world application has relied on informal handoffs, tacit know‑how, and opaque processes and is littered with failed attempts and learnings buried in lab notebooks. The industry has quietly accepted this agonizingly slow process and inability to broadly reproduce and distribute new discoveries.

We’re determined to support a new paradigm that ensures new scientific innovations can be more robust, interpretable, and transferable.

We are building processes and cutting‑edge AI tools that aim to illuminate and streamline these critical transfers of knowledge. We learn as much, if not much more, from failures as successes, transforming experimental science into robust, scalable solutions.

We aren’t building AI to replace human ingenuity, but rather AI that can be a better partner in understanding and translating that ingenuity to new domains. We seek to build tools that help scientists spend less time documenting and debugging and more time doing the exploration they love. To be clear: we believe this will become critical infrastructure. Without solving for scale and knowledge distribution, the discoveries that will be enabled by great scientists at the bench and generative AI designs will fail to reach their end customer: us.

Your impact starts here. We enthusiastically invite applicants across a broad range of expertise and experience to apply. Maybe you have expertise at the bench and have seen reproducibility challenges first hand, maybe you’re one of the first people to read a new AI benchmarking paper that comes out, or maybe you have a knack for quickly developing and deploying solutions that make mundane tasks easier.

Each candidate that joins the team will be a member of the technical staff, and compensation will depend on experience and responsibility. Our team works across AI and biotech innovation, so even if you apply for one role, we may consider you for others as well.

We think in‑person collaboration is invaluable, but recognize that talent is broadly distributed, and are open to flexible working arrangements. We’re based in Cambridge, MA, but invite applications from all geographies.

• Opportunity to join a creative and mission‑oriented founding team and build with us, from the ground up
• We have a bias for action and are obsessed with solving our customers’ real problems
• We also love bold and audacious science, enabling moonshots, and are motivated to work with and across the entire ecosystem, enabling new and better futures for ourselves and our families
• We also offer a full HR comp stack to include competitive salary, equity, and benefits

• Design and implement novel machine learning approaches—especially in pre‑training, multimodal learning, and reinforcement learning—to make experimental science more reproducible, interpretable, and transferable at scale.
• Design and develop AI models that can learn from experimental protocols, lab execution date, and failure modes to suggest, simulate, or guide future experiments.
• Develop and implement data collection strategies—including computer vision, audio, and sensor‑driven interfaces—to capture tacit knowledge embedded in physical lab workflows and human decision‑making.
• Become customer obsessed. Initiate, support, and lead program execution with external partners and collaborators to capture, document, and deliver results.
• Develop and test internal and external benchmarks, validation strategies, and frameworks for testing reproducibility, robustness, and scientific utility.
• Work alongside the biology and operations teams to identify automation opportunities, uncover latent structures in messy lab data, capture tacit knowledge, and improve experiment documentation…