Typical scientific AI
Correlation
- Predicts a property from past data
- Cannot say which ingredient, temperature, or process step is responsible
- A high confidence score is still a pattern, not a reason to intervene
Hong Kong
See what actually causes the result.
Asha Causal AI works with chemists, materials scientists, and medical researchers who already have data, and want a clearer why before they run the next experiment.
Prediction is useful. Cause is what you can act on.
Most models are good at spotting patterns. They are less good at telling a formulation scientist which ingredient, temperature, or process step is responsible.
We help you read experiments and simulations for cause and effect, so a recommendation is something a chemist or clinician can stand behind.
Most AI finds patterns.
We look for cause.
If two things move together, ordinary AI treats that as insight. In a lab, that is often a coincidence. Causal AI asks a different question: if you change this, does the outcome actually change?
Asha Causal AI
Causation
- Tests what would happen if you changed a variable
- Separates real drivers from side-effects that only look related
- Returns a model a chemist or clinician can read, challenge, and use
How it works
-
01
Read the scientific system
We start from your experiments and simulations, in the language of chemistry and process, not a generic spreadsheet.
-
02
Test cause against coincidence
We check which variables still matter if you imagine changing them, and which only appeared related.
-
03
Hand back a usable why
You get an account of what to change next, and why that change should work, rather than a black-box prediction.
Where this is useful
Materials and chemistry
When a battery, catalyst, or polymer does not behave as expected, we help you see which chemistry and process choices are really driving performance.
Medicine and formulation
In traditional medicine and biochemical work, we help separate the components and conditions that matter from those that only look related.
Everyday R&D decisions
If you already run experiments or simulations, we help you spend the next cycle on the lever that actually moves the outcome.
People
A small Hong Kong team spanning computational chemistry, scientific strategy, and engineering.
Prof. James P. Lewis
Founder and Chief Scientist
Computational physics and the FIREBALL quantum chemistry code.
Andrea Grippi
Senior Tech Officer
Scientific software, modeling, and infrastructure.
Prof. Joon Nak Choi
Advisor
HKUST Adjunct Associate Professor, entrepreneur, and AI strategist.
If this sounds useful, write to us.
Tell us what you are trying to understand in the lab. We work with research teams and design partners in chemistry, materials, and medicine.
info@ashacausal.ai