Palo Alto, CA — September 17, 2025 — Leads & Copy — Biostate AI has launched K-Dense Beta, a multi-agent AI research system designed to accelerate biological research and eliminate hallucinations in generative AI models. According to Biostate AI, K-Dense can compress research cycles from years to days. The company announced that K-Dense made a scientific breakthrough in longevity research during testing; the results will be published in a peer-reviewed journal later this year.
K-Dense coordinates specialized agents to plan experiments, review literature, design analyses, execute code in secure sandboxes, and generate publication-ready reports. The system cross-checks references against external databases and adds feedback loops.
Ashwin Gopinath, Co-founder and Chief Technology Officer of Biostate AI, stated that the company has created an AI scientist that can work 24/7, accelerating discovery while maintaining scientific standards.
K-Dense integrates access to bioinformatics pipelines, tools like Google’s AlphaFold, curated databases, and multiple specialized Large Language Models (LLMs) like MedGemma. It can also connect to any tool available through Model Context Protocol (MCP).
Professor David Sinclair, Co-Director of the Paul F. Glenn Center for Biology of Aging Research at Harvard Medical School, validated K-Dense’s capabilities. K-Dense analyzed the ArchS4 dataset, selecting 60,000 high-quality samples and focusing on 5,000 genes. The analysis revealed that different sets of RNA transcripts become important predictors at different points in life, showing that aging is not uniform but a sequence of biological programs.
According to Sinclair, K-Dense enabled the completion of an entire research study in just a few weeks, pointing to markers and pathways that warrant deeper study and helping build a unified AI model for predicting biological age.
The findings have been submitted for peer review and are available as a bioRxiv preprint. Biostate AI is validating K-Dense with academic institutions, biotechnology startups, and major pharmaceutical companies.
Bikram Singh Bedi, Vice President, Google Cloud Asia Pacific, said that Biostate’s implementation with Gemini 2.5 Pro showcases the model’s transformative potential for complex scientific challenges.
Built on Google Cloud’s Gemini 2.5 Pro, K-Dense achieved 29.2 percent accuracy on BixBench, surpassing models including GPT-5 (22.9 percent), GPT-4o (18 percent), and Claude 3.5 Sonnet (18 percent).
Since closing a $12 million Series A led by Accel, Biostate AI has expanded its capabilities and initiated clinical collaborations with MGH in the US, as well as partners in China and India.
K-Dense Beta is available to select design partners, with broader availability planned for later this year.
Contact:
Ashwin Gopinath, Co-founder and Chief Technology Officer of Biostate AI.