Younet.ai Highlights Researgency.ai Collaboration with Kala Bio (KALA) as New "AutoResearch" Milestone Signals the Rise of Autonomous Research Agents

Open-source breakthrough demonstrates AI agents can run 100+ experiments overnight-reinforcing the agentic R&D paradigm Researgency.ai is bringing to life sciences with Kala Bio (KALA)

Toronto, Ontario--(Newsfile Corp. - March 11, 2026) - Younet.ai, a Toronto-based AI technology company specializing in enterprise-grade private LLM solutions, today announced that the recent release of AutoResearch by AI pioneer Andrej Karpathy provides strong external validation for the autonomous research-agent approach underpinning Researgency.ai, Younet's agentic research platform being developed in collaboration with Kala Bio, Inc. (NASDAQ: KALA).

AutoResearch is an open-source tool that enables AI agents to autonomously conduct 100+ machine-learning experiments overnight on a single GPU, demonstrating a step-change in research velocity via continuous agentic iteration. Younet.ai believes the same "overnight research loop" can reshape life-sciences planning and decision-making through Researgency.ai, which is being developed with Kala Bio to apply autonomous research principles to biotech R&D workflows, including scenario simulation and protocol optimization.

"What AutoResearch demonstrates for machine-learning experimentation is the same transformation Researgency.ai is engineered to deliver for biotech research planning—continuous agentic iteration while teams sleep," said Alex Kapralov, CEO of Younet.ai. "With Researgency.ai and our collaboration with Kala Bio (NASDAQ: KALA), the goal is to help life-sciences teams wake up to better study designs, stronger protocol options, and faster evidence-driven decisions."

AutoResearch: A Public Blueprint for Autonomous Discovery

AutoResearch has been described as a compact framework that changes how experimentation can be conducted by enabling an agent to:

• Modify code, run experiments, and evaluate results autonomously
• Continuously iterate (~12 experiments per hour; 100+ overnight)
• Self-evaluate and keep improvements while discarding failures
• Operate efficiently on a single GPU, lowering the barrier to high-frequency experimentation

Karpathy has described the aspiration as enabling faster research progress "indefinitely" and without ongoing human involvement—an approach Younet.ai views as aligned with the agentic future emerging across R&D and knowledge work.

Schreibe Deinen Kommentar