ReactWise can now search the literature for you, extract what’s relevant across multiple papers, and combine it with MemoryBO to propose your starting points.

ReactWise can now search the literature for you, extract what’s relevant across multiple papers, and combine it with MemoryBO to propose your starting points.

Every campaign starts with a guess. Someone decides what solvent, what base, what temperature goes on the plate - from experience, a few half-remembered papers, and whatever the group ran last time. That guess sets the boundaries of everything the campaign can discover.

Our literature search feature finds the papers that are actually relevant to your transformation and pulls out the passages that bear on it. Not a reading list. The conditions, substrates, and observations that matter for what you’re about to run.

Then it shows its work. Which papers, which passages, why it landed where it did. Every recommendation traces back to a source you can open and overrule.

The part we thought about hardest is bias. Published chemistry is a record of what worked, not of what was tried. A recommendation built on that alone will be confidently wrong about where a process stops working.

So literature isn’t the only input. It’s combined with our own descriptor databases and data from our lab, including the negative results that never reach a paper.

What it replaces is an afternoon of skimming and writing down conditions from four papers that nearly match.

Step zero used to be guesswork. Now it’s a starting point you can defend.

‍