Biostate AI has launched K-Dense Beta, an artificial intelligence system designed to condense biological research cycles from years to days, with its efficacy demonstrated in longevity research in collaboration with Harvard Medical School.
Scientific and pharmaceutical research is generating increasing amounts of data, with industry figures suggesting biotech data is doubling every 7 to 8 months. Despite the growth, it is estimated that 90% of this information remains unused, leaving undiscovered insights buried within existing datasets.
AI research system
K-Dense Beta operates as a multi-agent AI system that simulates collaborative scientific teams. The platform coordinates specialised agents responsible for experiment planning, literature review, data analysis, secure coding, and compiling publication-grade reports. Decisions are cross-verified by these independent agents to minimise so-called hallucinations, a known challenge in generative AI applications, and to build full traceability of each system action.
According to Ashwin Gopinath, Co-founder and Chief Technology Officer of Biostate AI, this new approach is intended to address the current challenges in data evaluation within the scientific community.
"There is a crisis in science right now, where we have too much data and not enough resources to evaluate it. We have created an AI scientist that can work 24/7, dramatically accelerating discovery while maintaining rigorous scientific standards," said Ashwin Gopinath, Co-founder and Chief Technology Officer of Biostate AI.
The platform can integrate with standard bioinformatics tools, curated databases, and large language models such as MedGemma, and can connect to further software via the Model Context Protocol (MCP).
Longevity research validation
K-Dense's performance has been tested through a collaboration with Professor David Sinclair, Co-Director of the Paul F. Glenn Centre for Biology of Ageing Research at Harvard Medical School. The AI system was tasked with generating a transcriptomic aging clock by analysing the ArchS4 dataset, which contains over 600,000 transcriptomic profiles.
During the project, K-Dense selected 60,000 high-quality samples and identified 5,000 genes most relevant from nearly 50,000 available. The system's analysis determined that RNA transcripts become predictive of biological age at different times throughout life, supporting the view that aging is not uniform but progresses through distinct biological programmes.
Professor Sinclair noted that the research cycle, usually requiring months or even years, was completed within weeks.
"K-Dense enabled us to complete an entire research study in just a few weeks, work that typically requires months or years of expert analysis. It pointed us to markers and pathways that warrant deeper study and helped us build a unified AI model for predicting biological age. Importantly, it also provided a measure of how reliable those predictions are, which is critical for scientific applications and has not been available in prior AI approaches," said Professor David Sinclair at Harvard Medical School.
The findings from this study have been submitted for peer review and have been shared as a preprint on the bioRxiv platform. Biostate AI is now working with dozens of design partners, spanning academic research organisations, biotech startups, and major pharmaceutical companies, to validate K-Dense's further capabilities.
Industry benchmarking and collaboration
K-Dense Beta has achieved a 29.2 percent accuracy score on BixBench, a commonly used benchmark for bioinformatics AI, outperforming established models such as GPT-5 and Claude 3.5 Sonnet. The company attributes this performance in part to its implementation of Google Cloud's Gemini 2.5 Pro technology.
"Biostate's implementation with Gemini 2.5 Pro showcases our model's transformative potential for complex scientific challenges. Their multi-agent approach demonstrates how intelligent coordination of advanced language models can accelerate genuine scientific discovery," said Bikram Singh Bedi, Vice President, Google Cloud Asia Pacific.
Since closing a USD $12 million Series A funding round led by Accel earlier this year, Biostate AI has expanded its development activities, completed the K-Dense system, and initiated partnerships with hospitals and research centres in the United States, China, and India. Other investors include Dario Amodei of Anthropic, Emily Leproust of Twist Bioscience, and Mike Schnall-Levin of 10x Genomics.
This launch represents the latest effort by technology companies and research institutions to address underutilised data in scientific fields, by automating and accelerating analysis through modular agent-based artificial intelligence systems.
K-Dense Beta is being made available currently to select design partners, with wider rollout plans expected later in the year.
