SynSilico launches Agentic INNOptimizer for experimental design

SynSilico has released Agentic INNOptimizer, a web-based optimization platform that uses Bayesian Optimization and a user-controlled agent to cut the number of experiments needed for R&D. The company says the tool can reduce experiments by up to 70% in complex optimization tasks and is available now with a free trial at optimizer.synsilico.com.

Why it matters: - Agentic INNOptimizer is aimed at reducing the cost, time and trial-and-error burden of R&D across chemistry, materials, pharmaceuticals, food technology and other fields. - SynSilico says the platform can cut experiment counts by up to 70% versus standard design-of-experiments tools in complex optimization work. - The product is designed for teams running expensive or resource-intensive experiments, where each iteration can slow development timelines.

What happened: - SynSilico launched Agentic INNOptimizer, a web-based optimization platform for experimental design. - The release came from Geleen, Limburg, Netherlands, on August 25, 2026. - The platform is accessible at optimizer.synsilico.com. - SynSilico also announced a 20% discounted early-adopter price for the first two months, through Nov. 30, 2026. - A free trial version is available for immediate use with no installation required.

The details: - Agentic INNOptimizer uses Bayesian Optimization and Optibot, a user-controlled agent, to configure optimization tasks dynamically. - The platform automatically proposes BO configurations and can adjust settings during optimization cycles when it detects improvement potential. - The system checks for potential outliers and aims to reduce redundant experimentation. - The tool includes visualization and analytics for design-space exploration, response-surface maps, Pareto frontiers and convergence behavior. - The platform supports multivariate experimental design, seeding strategies, constraint setup, sequential learning with uncertainty quantification, and agent-supervised exploration and exploitation. - The software includes an API for integration with external software, laboratory automation systems and digital co-pilots. - SynSilico says the API makes the tool suitable for self-driving lab applications and closed-loop experimental workflows. - The company says customers build a private dataset of experiments and outcomes as they use the platform. - SynSilico offers local implementation, data strategy consulting and self-driving lab integration services.

Between the lines: - The launch reflects a broader push to move experimental R&D from static statistical planning toward adaptive, software-guided decision-making. - The agentic layer is meant to reduce the need for expert users to manually tune Bayesian Optimization settings. - The product also positions SynSilico in the growing market for digital labs and automated experimentation. - Meta published a paper on agentic Bayesian optimization on Aug. 8, which SynSilico cited as context for the approach.

What's next: - Researchers and engineers can test the platform now through the free trial at optimizer.synsilico.com. - SynSilico expects the tool to be used in process optimization, formulation work and QbD studies across multiple industries. - As users accumulate data in private environments, SynSilico is likely to lean on consulting and integration services to expand adoption.

The bottom line: - SynSilico is betting that combining Bayesian Optimization with a user-controlled agent will make experimentation faster, cheaper and more adaptive for industrial R&D.