Designed, not discovered by chance
No compound reaches synthesis until every discipline in the room has had a chance to reject it.
Our story
Chemtelliq was founded in Seoul in 2024 as a deliberately multidisciplinary team. A molecule has to answer to more than one field — how it behaves physically, what it does inside a cell, whether it could ever reach a patient — and those questions are rarely asked at the same time.
So we put them at the same table. Physics, chemistry, biology, pharmacy, medicine and mathematics are represented in the room where a candidate is designed, not consulted afterwards. It makes the design conversation slower and considerably more rigorous, which is the trade we wanted.
The result is a slower path to the first synthesis, and a much shorter list of things we turn out to have been wrong about.
What we believe
- Data quality beats data volume. A clean, internally consistent assay set outperforms a large noisy one.
- Negative results are training data. A compound that failed tells a model something no successful compound can. We keep, curate and train on the failures with the same care as the hits — a model shown only what worked has never learned where the edges are.
- Developability is a design constraint. ADME and safety liabilities are cheaper to design out than to fix.
- Publish and be challenged. We put our methods in front of the field.
Milestones
-
2024
Company founded
Incorporated in Seocho-gu, Seoul.
We are hiring
If you want short feedback loops and real ownership of a program, we would like to hear from you.