The decisions that shape a synthesis program belong on day one.

A SynovAI team ran a prospective retrosynthesis on orforglipron, a major oral GLP-1 receptor agonist, asking whether the published route could be shortened without losing yield.

Within a week the platform surfaced a synthesis that improved on the published pathway. A month later, the company holding the asset disclosed an updated route working through the same disconnections. They had been on the program for roughly two years.

Two graduate students used the platform to plan, run, and characterize an improved synthesis for a target that an earlier 20-person team had spent two years on.

The full loop, from entering the molecule to confirming the product in the lab, took two weeks. The platform did the planning; the students did the bench work.

A chemist had spent two weeks searching for a synthetic route to a tertiary amine using the standard discovery tools in the industry. They found one viable pathway.

Running the same target through SynovAI took ten minutes. The platform returned the same route they had found, plus two additional routes the existing tooling had missed. One of the new routes used cheaper, more accessible reactants.