Briefly Bio, a London-based startup, has secured $1.2 million in funding with the aim of addressing the reproducibility crisis in scientific experiments. This crisis refers to the inability to reproduce or replicate many scientific experiments, which has raised concerns among researchers. To tackle this problem, Briefly Bio has developed a platform that utilizes large language models (LLM), similar to those used in leading AI products like ChatGPT, to convert complex lab documentation into a consistent, structured format. The goal is to make it easier for scientists and engineers to use and build upon existing research.
Currently, scientists document their research in various ways, leading to ambiguity and the loss of critical details. This lack of consistency in documentation hampers collaboration and makes it difficult to reproduce experiments, resulting in an estimated annual cost of over $50 billion to the industry. Briefly Bio’s platform addresses this issue by converting natural language scientific protocols into a structured format, providing step-by-step information for reproducing or building upon experiments.
The founders of Briefly Bio, Harry Rickerby, Katya Putintseva, and Staffan Piledahl, have firsthand experience with the challenges posed by inconsistent documentation. Through their respective careers in academia, data science, and automation engineering, they realized the common root cause of many struggles was the lack of consistent documentation of lab work. This led them to create Briefly Bio as a solution.
Briefly Bio’s tool utilizes generative AI to extract key information from plain text descriptions and categorize them into different processes, actions, explanations, and parameters. This structured representation is then transformed into a visual format that is easier to understand than traditional text descriptions. The platform also includes an AI copilot that can spot errors and identify parameters related to the lab work. It enriches the hierarchical representation of the method and generates missing parameters within seconds.
In addition to converting existing scientific descriptions into a structured format, Briefly Bio’s platform includes a workspace where teams can reuse the generated data in their experiments. Users can mark each step as complete or incomplete, add calculations and text, and create a visual representation of their experiment.
While Briefly Bio is still in its early stages, the company has already started generating revenue from its first customers. The platform caters to wet lab scientists in early-stage research and development, as well as those working in laboratory automation. Looking ahead, Briefly Bio aims to create a public version of its platform, similar to GitHub, where scientists can share experiments and protocols. This would enable scientists to easily discover reproducible methodologies that they can adapt for their own labs.
In summary, Briefly Bio’s platform addresses the reproducibility crisis in scientific experiments by providing a consistent and structured format for lab documentation. By utilizing AI and creating a shared language for data understanding, the platform aims to improve collaboration and reduce the cost of failed reproductions. With plans to create a public version of the platform, Briefly Bio aims to revolutionize the way scientists share and collaborate on experiments, much like GitHub did for open-source software development.