Bluenote has officially announced $10M in financing for its generative AI platform, which is designed to serve life sciences companies.
Led by Lux Capital, the round saw further participation coming from the likes of Elad Gil, Anthropic & Menlo Ventures Anthology Fund, McKesson Ventures, Avichal Garg/Electric Capital, Moxxie Ventures, Carbon Silicon Ventures, and leaders in AI and life sciences – Othman Laraki (CEO Color Health), and Fidji Simo (CEO Instacart, Co-founder Metrodora Institute, Not just that, OpenAI Board Mike Nohaile (CEO Prellis Biologics, previously Amgen & Novartis Executive), Kristen Fortney (CEO BioAge), Eric Morgen (COO BioAge), Qasar Younis (CEO Applied Intuition), Linus Upson (Verily), and Jeffrey Low (Life sciences investor) also joined the mix.
To understand the significance of such a development, we must take into account how companies operating in the life sciences industry are heavily regulated, and therefore, they are required to invest significant time and resources to fulfill their regulatory and compliance obligations.
Now, while the regulatory framework is in place to ensure patient safety, clinical efficacy, and consistent product quality, the work required to produce compliance documentation is time-consuming and largely manual. You see, once a trial is completed, companies have to typically spend 8-9 months preparing regulatory submissions. As a result, thousands of scientists, engineers, and development professionals spanning multiple functions, including research, clinical, manufacturing, regulatory and quality, dedicate significant time to produce thousands of pages for regulatory filings and manage complex interdependencies between protocols, technical reports, and compliance requirements.
Fortunately for them, Bluenote’s proprietary platform arrives on the scene with an ability to streamline regulatory workflows and bring their breakthroughs to patients sooner. The solution’s most widely used applications, at the moment, focus on producing regulatory filings, technical reports, protocols, Standard Operating Procedures (SOPs), validation reports, risk analyses, manufacturing documentation and more.
Another detail worth a mention here is rooted in the fact that Bluenote has deployed more than 15 applications for scientists, engineers, quality, regulatory and manufacturing teams. It has even developed fine-tuned models that are 90% preferred over off-the-shelf models.
“Within the healthcare industry, we face extensive regulatory requirements and documentation that must adhere to numerous guidelines and regional laws. At Guardant Health, we have been pleased with Bluenote’s generative AI capabilities to streamline these time-intensive and complex tasks—ranging from scientific research papers, study reports, and software development documentation critical to regulatory filings,” said Kenny Speer, Vice President, Bioinformatics and Software Engineering at Guardant Health. “Bluenote’s technology not only helps the accuracy, reliability, and timeliness of our documentation, but also flags areas that need human-in-the-loop review.”
Talk about the given platform on a slightly deeper level, we begin from the accuracy and verifiability of its AI outputs. In essence, Bluenote’s product is engineered to include only factual details from traceable primary sources, combining guardrails with custom Large Language Models (LLMs). Furthermore, it inserts call-to-action placeholders for human reviewers to provide details where greater context is needed.
Next up, we must get into its promise to provide customized outputs that are generated on the back of proprietary datasets, as well as connectivity to data lakes & a multi-model approach. This translates to how Bluenote developed domain-optimized Retrieval-Augmented Generation (RAG) which ingests data from various data lakes and third party applications, accurately processes the datasets, including industry-specific edge cases, and indexes them to support unique application requirements.
Moving on, the platform in question also combines the best models from Anthropic, OpenAI, Google, and its own fine-tuned models based on proprietary datasets. Complementing the same would be its bid to continuously evaluate and update model configurations to ensure that customers are among the first to benefit from the latest advances.
Bluenote also delivers at your disposal a single secure platform which supports multiple workflows and functions. Basically, by leveraging a single platform to run AI applications, the same proprietary knowledge base is used for multiple applications. Besides that, outputs from one workflow become inputs into another, enabling the propagation of changes. On top of that, customers are also able to consolidate their proprietary data within a single, secure environment for simplicity.
