Faster scale-up and tech transfer with AI-enabled digital twins
Predict process performance at a new scale or site before you run it, with fewer engineering runs and consistent CQAs.
Run your scale-up DoE virtually. Reduce physical trials by up to 50%.
Test scale-up strategies and operating conditions on a digital twin of the target equipment. Then run only the batches you need to confirm the prediction.
Right-first-time scale-up
with predictive physics-AI digital twins.
Simulate scale-up strategies, see predicted CQAs and equipment limits, and analyze the results with your team before the first batch is scheduled.
Reduce site-to-site variability. Hit the same CQAs at every site.
Basetwo shows where sites differ and why, and recommends site-specific process design and control strategies for the receiving site or CDMO.
How it works: A digital twin built on your data, equipment and process context.
Bring in historian, MES and LIMS data with equipment specs and SOPs. The agent aligns batches and flags gaps before modeling.
Start from a validated unit-operation template or your own model. Mechanistic structure plus machine learning, checked on held-out batches.
Run virtual experiments across scales and sites, each with a confidence interval, and share results as an app your team and receiving sites can use.
AI process agents for scale-up and tech transfer.
Use AI process agents across the whole workflow:
Proven in production at leading manufacturers. Explore the case studies.
Map the receiving site's equipment line by line and predict the first batch there before it is run.
Choose the scale-up criterion that keeps droplet size and blend time in spec at production scale.
Set perfusion and cell-bleed rates virtually before the first 2,000 L perfusion run.
Traditional scale-up and tech transfer vs. Basetwo.
Traditionally, each new scale or site needs its own studies and engineering runs, adjustments are estimated in offline spreadsheets, and problems are found in the plant. Every run costs material, equipment time and weeks of schedule. Basetwo runs those studies in simulation first, so plant runs confirm a prediction instead of searching for one.
Biologics tech transfer from a sending site to a CMO.
A global biopharma manufacturer replaced spreadsheet-based tech transfer analysis with site-specific digital twins, and simulated gassing and agitation adjustments so the receiving site matched the sending site's oxygen uptake rate.
Questions scale-up and tech transfer teams ask
Typically a few. Kinetic parameters carry over from earlier scales, and scale-dependent terms such as mass transfer, mixing time or heat removal are fit from your previous and historical scale-ups, so new runs confirm the prediction rather than build it.
That's the usual case. The model separates process kinetics from vessel physics, so a new geometry, impeller or jacket changes the vessel parameters, not the whole model. Anything the model can't resolve from the equipment specs is flagged with the run that would close it.
Every prediction carries a confidence interval.
Predict your first batch at scale.
