Coursera Spring Sale
40% Off Coursera Plus Annual!
Grab it
Explore the engineering challenges and solutions required to build AI Co-scientists that serve as true collaborative partners in scientific research through this 19-minute conference talk. Discover how to architect robust systems that bridge powerful AI reasoning with the dynamic reality of physical experiments, focusing on scalable multi-agent architectures that leverage foundation models like Gemini for complex reasoning and hypothesis refinement. Learn about the core engineering challenges of integrating diverse, real-time empirical data streams including visual data from cameras, quantitative sensor readings, positional feedback from actuators, and instrument outputs directly into AI reasoning loops. Examine concrete technical examples across chemistry with adaptive reaction monitoring, robotics featuring vision-guided assembly using SO Arm 100 and LeRobot library, and synthetic biology through real-time bacterial growth monitoring and interpretation. Understand engineering strategies for handling data heterogeneity, latency, and noise while enabling AI systems to interpret, correlate, and act upon live experimental feedback. Gain insights into how thoughtful engineering of AI Co-scientists can democratize access to advanced scientific capabilities, presented by Stefania Druga, former Research Scientist at Google DeepMind and independent researcher specializing in multimodal AI applications and Large Language Models.
