We offer both project-based CRO services and hourly consulting services

Discover our advanced AI-driven drug design services, customized to align with your R&D needs across diverse modalities. Benefit from quick outcomes, with computational outputs delivered in just one week, and a full package – encompassing computational analyses, molecule production, and wet lab testing – completed within a month. 

SentinusAI® Protein Design


Protein-Protein Binding Affinity Maturation

Therapeutic Antibody Modification

Our affinity maturation engine attempts to utilize AI techniques to perform in-silico therapeutic antibody maturations with better affinity improvement and least experiment cost.

Fusion Protein Engineering

We utilize sequence-based modeling approach to study and predict the binding affinity between fusion proteins and peptides.

Target-Antibody Binding Ranking Prediction

This module takes antibody/antigen sequences provided by the user and provide a ranking list for all antibody-antigen combinations in terms of their corresponding binding affinity values (AI predicted binding scores).

Protein Humanization

Antibody Humanness

This model is designed to evaluate the humanness of antibody which potentially indicates the level of immunogenicity risk.

Antibody Humanization

Humanization is a popular method to reduce the immunogenicity risk. This module is used to create human-like antibody with lower immunogenicity risk based on a novel algorithm

Protein Off-target Prediction

Protein Off-target toxicity evaluation

Therapeutic antibodies/proteins are designed to have high specificity towards targets with least side effects that may be harmful to the human body. The Off-target engine helps to predict the risk of unwanted binding activity for large molecule drugs.

CarbonAI® Compound Design


Molecule Generation

This structure and rule-based design engine is used to generate novel compounds with desired R-groups and new cores for more optimized pharmacological properties and easier synthesis.

Lead Optimization Analog Expansion

This structure and rule-based design engine generates novel analogs of a desired scaffold for more optimized pharmacological properties. This enumeration opens endless possibilities for compound optimization, and multiple rounds lead to infinite combinations of novel analogs.


This design engine generates novel PROTAC degrader compounds based on the input structure-based parameters. These structures can be used by other modules for subsequent property prediction.


Target Binding And Selectivity Optimization

Protein Pocket Binding Affinity Prediction

The protein-pocket structure and ligand interaction is modeled for rapid prediction of the binding affinity between a protein and multiple screening compounds. This ultra-efficient virtual screening allows for screening of millions of compounds in a single day.

Protein Sequence Binding Affinity Prediction

All proteins consist of a primary sequence of 20 basic amino acids. This universal model of a protein sequence with a small molecule allows for an ultra-efficient virtual screening workflow for screening millions of compounds per day without the need of a protein structure.

Protein Pocket Detection

We provide an efficient and accurate platform to identify druggable and functionally relevant candidate pockets in protein structures

Compound Off-target Prediction

Avoiding off-target effects of a small molecule is a key hurdle to tackle for any drug discovery program. This engine helps predict the risk of unwanted binding affinity for different target families such as kinases, transporters, or receptors so cleaner small molecules can be further prioritized.


General Pharmacological Profile Optimization


The ADME (Absorption, Distribution, Metabolism, and Excretion) prediction module allows you to find the most optimized molecule using over 20 models covering each of the four domains to predict assay and chemical properties. This will give you an early conclusion regarding which compounds can be prioritized for screening.

Toxicity Prediction

Different toxicity domains such as cytotoxicity, carcinogenicity, and organ-injury can be predicted. This allows for prioritization of compounds for screening in different stages of drug-discovery, lowering the risk of a failed program due to undesirable toxicity properties.