AI-driven discovery for drugs and materials, combining computational models with large-scale automated experimentation.

Cambridge, MA
Joined April 2016
In Singapore for #SWITCH2026? XtalPi is on the main stages and hosting an #AIforScience partner event on Oct 28. • Dr. Lipeng Lai on how AI and TechBio are reshaping drug discovery. • Dr. Peiyu Zhang on what could drive the next inflection point in materials AI. We’ll also cover how AI, agents, and robotic labs are advancing drug discovery at our partner event. Registration👇
1
1
3
295
Interested in our partner event? Here are the details: Oct 28 | 15:30–18:00 SGT Sands Expo, Level 3, Cassia Main 1 Register here: gevme.com/switch2026_xtalpi?…
2
161
Our first siRNA PCC is here, developed internally in just 9 months with Kodexia™. The IgAN candidate’s IC50 was tens of times lower than clinical-stage references in vitro. In nonhuman primates, one dose suppressed the target protein for months, longer than either reference.🧵
1
1
4
419
Kodexia’s predicted activity ranking across the IgAN candidates was confirmed in vivo, helping us select the molecule to advance. Results from high-throughput experiments fed back into the models to guide successive rounds of design.
1
1
26
IgAN is the first of our six proprietary siRNA programs to reach PCC. We’re now advancing it toward clinical development and looking to work with pharma partners to bring more such therapies through discovery and development. prnewswire.com/news-releases… #AIforScience #siRNA #DrugDiscovery
1
50
Automated labs can run thousands of reactions. But every new product poses the same question: how much did we actually make? XtalPi is open-sourcing lcms2yield, an AI model integrated into our robotic labs that estimates yields without product-specific calibration curves. 🧵
2
5
310
In our JACS Au study, we evaluated lcms2yield across 178 reactions and found a mean absolute yield error of 4.74 percentage points. Predicting UV correction factors was 10⁵-10⁶x faster than quantum chemistry calculations, supporting high-throughput yield measurement.
1
2
35
We've made lcms2yield available on GitHub - along with the code, model weights, and datasets. We hope others working on automated chemistry find it useful and build on it. GitHub: github.com/xtalpi-inc/lcms2y… Paper: pubs.acs.org/doi/10.1021/jac… #AIforScience #OpenSource
3
43
Ailux, XtalPi’s AI-native biotech subsidiary, welcomes Lloyd S Miller, MD, PhD as Chief Medical Officer. With 25 years of experience in drug R&D, including leadership at Johnson & Johnson Innovative Medicine, Dr. Miller will lead clinical development of Ailux’s biologics pipeline toward first-in-human studies targeted for 2027.
2
4
378
1/100th the concentration. Comparable hair-growth scores to 5% minoxidil in a DHT-induced androgenetic alopecia mouse model. XtalPi’s hair-loss candidate has entered IND-enabling studies, targeting dormant follicle stem cells through a different biological pathway. 🧵1/6
3
1
9
917
Mouse efficacy is only one hurdle. We’re also evaluating dose response, safety margins, and suitability for long-term use. A low required concentration may help reduce irritation risk. Pharmacology, safety, and CMC studies are underway as the program enters IND-enabling. 5/6
1
2
67