Lab automation for faster ML-driven protein engineering
AI is changing how biologics are discovered. But that can pose a challenge for wet labs–from data requirements to speed of iteration. In this webinar, we will share some of the actual strategies we use to increase turnaround times while keeping data quality high.
Join Cradle bioengineers Michelle Vandeloo and Jack Cunha when they share their journey to faster protein rounds - from scaling from 48 variants to parallel 384-well rounds while cutting hands-on time up to 75%. You can also look forward to a hands-on discussion about how we designed our workcell to power our ML-driven protein engineering with our partner LAB Services.
Date & time
Join us live on Oct 14, 2025 , 16:00h CEST (10AM EDT)
We’ll discuss
Why wet labs are critical for ML-driven protein engineering
Upstream and downstream challenges in ML-driven protein engineering environments
Strategies to improve workflows beyond just lab instruments
Designing an ML-ready workcell
How we scaled throughput from 48 variants to parallel 384-well rounds
Case study: Fixing our purification process
Live Q&A
We'll share practical protocols, automation setups, and real troubleshooting examples from our Amsterdam lab.
Who is this for?
This webinar is most interesting for R&D leaders and experimental scientists across modalities - you’ll get a glimpse into our wet lab workflows and build an understanding of how we tackle lab workflows in an ML-driven protein engineering environment.
Please submit questions!
There will be a dedicated Q&A section. We would love to hear from you about your main challenges and questions our experts can answer during the webinar.
Speakers
Michelle Vandeloo (Research Associate, Cradle)
Jack Cunha (Lab Automation Engineer, Cradle)
Tim Verweij (Global Automation Expert, Lab Services)
Before you sign up…
Please do not use your gmail to sign up. We will review webinar signups. Please register with your work email address - we will review and approve attendees and non-work or .edu emails will be auto rejected.