Evogene reports a major Google Cloud collaboration milestone advancing agentic AI systems to improve early-stage small molecule discovery and optimisation workflows.

Evogene Ltd has announced a new milestone in its collaboration with Google Cloud, advancing its ChemPass AI platform (ChemPass AI) toward more autonomous systems for small molecule discovery and optimisation.

The update centres on the integration of agentic computational systems into ChemPass AI, designed to support earlier identification of key biological and clinical success criteria during drug discovery. According to the company, the goal is to improve scalability, reduce attrition rates in development pipelines and increase the probability of success in small molecule programmes.

Evogene describes the approach as enabling earlier definition of Target Product Profile (TPP) requirements, which typically guide whether a molecule is likely to succeed in later development stages. The company argues that these criteria are often assessed too late in traditional workflows, contributing to inefficiencies and costly programme failures.

The enhanced platform is built using Google Cloud’s AI infrastructure, including Gemini models and the Gemini Enterprise Agent Platform. Evogene says these tools allow multiple autonomous AI agents to analyse structural, chemical, biological and clinical datasets simultaneously in order to prioritise compounds more effectively at the earliest stages of discovery.

The company said the system is designed to “enable earlier and deeper biological TPP profiling, with the goal of identifying the critical criteria a molecule must satisfy to progress successfully.”

It added that the integration of agentic workflows is intended to allow research teams to anticipate downstream development requirements earlier, reducing reliance on late-stage filtering and improving decision-making during candidate selection.

Boaz Maoz, managing director of Google Cloud Israel, said: “We are excited to see how Evogene is leveraging Google Cloud’s full AI stack to redefine the life sciences sector. By using our Gemini models and embarking on agentic AI, Evogene is positioned to innovate and significantly accelerate the discovery of new molecules, demonstrating the transformative power of AI in streamlining complex biological and chemical research.”

Ofer Haviv, president and chief executive officer of Evogene, said: “This milestone advances our vision to redefine molecular discovery and optimization through AI. Google Cloud’s technology provides a robust foundation for scaling these capabilities across our pipeline and future collaborations. By integrating advanced AI agents into ChemPass AI, we believe we can generate richer target profiles earlier in discovery. This is intended to allow us to anticipate downstream requirements and prioritize high-potential candidates before committing significant resources, with the goal of moving us closer to autonomous computational systems that predict key success requirements at the earliest stages.”

He added that the company believes the approach may help address complex protein targets that remain challenging in traditional drug discovery pipelines.

The collaboration highlights a broader trend in drug discovery towards the use of cloud-based AI systems and multi-agent architectures to streamline early-stage decision-making. By shifting parts of the target selection and profiling process earlier in development, companies are aiming to reduce late-stage failure rates, which remain a major cost driver in both pharmaceutical and agricultural small molecule development.

While the announcement outlines potential efficiency and scalability benefits, it remains a platform development milestone rather than a report of clinical or preclinical trial outcomes. The real-world impact of the system will depend on its performance in future discovery programmes and validated pipeline outputs.