FAQs

Question 1

What is ReactWise?

Accepted Answer

ReactWise is an AI copilot for chemical process optimization. Our cloud-based platform uses data-driven optimization and machine learning to help R&D teams in pharma, fine chemicals, biotech, and beyond run smarter experiments, reduce wasted runs, and reach optimal conditions faster. ReactWise is backed by Y Combinator and Innovate UK, and is trusted by organizations including Pfizer, Amgen, Biogen, and the University of Cambridge.

Question 2

Who is ReactWise for?

Accepted Answer

ReactWise is designed for bench chemists, process chemists and engineers, automation scientists, and biotechnologists who want to accelerate process development using data-driven optimization. Managers and team leaders use the platform to track project progress and improve data discipline across teams.

Question 3

What industries does ReactWise supports?

Accepted Answer

ReactWise supports industries including pharma, fine chemicals, biotech, and related R&D sectors. For more details, visit our industries page and case studies.

Question 4

What are the benefits of using ReactWise?

Accepted Answer

Benefits include faster development, better outcomes, lower costs, and no coding required.

Question 5

What is ReactWise's core offer?

Accepted Answer

ReactWise offers four core products: Optimization Platform, HTE Plate Designer, Solvent Selector, and Kinetic Modeling. For more information, visit our products page.

Question 6

How does ReactWise's software work, and how is it different from Design of Experiments?

Accepted Answer

Design of Experiments (DoE) requires you to define a fixed experimental plan upfront and typically treats each campaign in isolation. ReactWise uses Bayesian optimization: a data-driven model learns from every experiment, quantifies uncertainty across the design space, and recommends the most informative next conditions to test.

This means fewer experiments to reach optimal conditions, because the algorithm balances exploring untested regions with exploiting what already looks promising. For example, a project with three continuous and two categorical variables might require 30–50+ experiments under DoE, while Bayesian optimization can begin identifying optimal conditions from as few as 10 initial experiments, iterating intelligently from there. With transfer learning (MemoryBO®), the model can also transfer knowledge from previous related campaigns, so you are not starting from scratch every time.

Learn more on our optimization platform page.

Question 7

What is MemoryBO® and how does transfer learning work?

Accepted Answer

MemoryBO® is ReactWise's proprietary multi-task Bayesian optimization algorithm. It models related reactions and learns from historical campaigns relevant to your current target. Before a single new experiment is run, the model already arrives at an informed starting point.

Question 8

Can ReactWise handle multi-objective optimization?

Accepted Answer

Yes. ReactWise supports simultaneous optimization of multiple competing objectives, such as maximizing yield while minimizing impurity or cost, and can assign weights to these objectives. Inputs can also be treated as objectives, for example, looking to minimize the loading of an expensive catalyst.

Question 9

Is ReactWise chemistry informed?

Accepted Answer

ReactWise combines data-driven Bayesian optimization with built-in chemical intelligence. We do this in two ways:

Transfer learning from large in-house or historic datasets to train MemoryBO® models that are informed by similar campaigns.
Descriptor databases for acids, bases, solvents, ligands, and catalysts (e.g. steric and electronic properties for ligands, pKa for acids, or polarity for solvents etc.) that allow the software to model these as chemicals.

‍

Question 10

How does ReactWise ensure the accuracy of its predictions?

Accepted Answer

We validate models using extensive real-world industrial data, and our platform is built on proprietary datasets - over 25,000 HTE data points across reaction classes such as Suzuki, amide, and Buchwald-Hartwig couplings - which are used to pre-train reactivity models. Every prediction includes a confidence range so you can assess reliability before committing to experiments.

Question 11

How can I get started with ReactWise?

Accepted Answer

Getting started is straightforward. Book a demo with our team to see the platform in action. After that, we guide you through onboarding: you can set up your first optimization in as little as five minutes by defining your parameters, objectives, and constraints. We support you with regular check-ins throughout your first projects.

‍

Question 12

How many experiments do I need to get started?

Accepted Answer

You can start an optimization campaign from scratch with as few as a handful of initialization experiments. ReactWise's Bayesian optimization is designed to work with small datasets and improve with each new data point. If you have historical data from your ELN or previous campaigns, MemoryBO® can use transfer learning to give the model an informed starting point, reducing the total number of experiments needed.

Question 13

What kind of support does ReactWise offer?

Accepted Answer

We provide hands-on onboarding, technical documentation, and ongoing support. For immediate help, our AI assistant Vera is available in-platform. For further support, our technical team is available via email and can offer next-day calls. We also have regular check-ins to ensure you are getting the most from the platform.

Question 14

Is ReactWise easy to integrate with existing processes?

Accepted Answer

Yes. ReactWise is designed to fit into your existing workflows without disruption. You can upload data from spreadsheets, connect your ELN for direct data exchange, or use our API to build custom integrations. The platform works with the data you already have, so there is no need to change how your lab operates before getting started.

Question 15

Can ReactWise integrate with automated hardware?

Accepted Answer

Yes. ReactWise has a fully programmable API that can connect to any automated hardware for closed-loop, real-time optimization. We have integrated with automated liquid handlers, PAT instruments, and flow chemistry equipment. Our team is happy to discuss custom integration requirements. Full API documentation is available for clients.

Question 16

Can ReactWise integrate with our current software?

Accepted Answer

Yes. ReactWise can integrate with current software systems. For more information, visit our integration page.

Question 17

Can ReactWise be deployed on-prem?

Accepted Answer

Yes, we offer on-premises deployment and can work with your IT team to design a solution that meets your infrastructure and compliance requirements.

Question 18

How secure is my data with ReactWise?

Accepted Answer

ReactWise takes data security seriously. For detailed information about our security practices, please visit our Trust Center.

Question 19

Does ReactWise have lab capabilities?

Accepted Answer

Yes. ReactWise operates a fully equipped, automated wet lab in Cambridge with an in-house chemistry team. The lab serves two functions: generating large-scale HTE datasets to pre-train the reactivity models available on our platform (over 25,000 data points generated across Suzuki, amide, and Buchwald-Hartwig couplings to date), and running experimental campaigns on demand for clients who need structured data for a specific reaction class.

All datasets are delivered as machine-readable, FAIR-compliant files in ORD format. Learn more about our lab facilities and HTE Data on Demand service.

Question 20

What types of reactions and processes has ReactWise been used for?

Accepted Answer

ReactWise has been used for various types of reactions and processes. For specific examples, please visit our case studies page.

Question 21

Where can I find more information?

Accepted Answer

You can find more information in our white paper, publications, or by contacting us through our contact form.

About Reactwise

Our products & technology

Getting started

Integration & deployment

Data security & privacy

Lab services

Miscellaneous