AEMSFormulation engineering
AEMS develops formulations to reduce viscosity at high concentrations. We combine high-throughput experiments, robotic automation, and machine learning to explore established excipients for smaller injection volumes and subcutaneous development.
Why viscosity matters
A higher protein concentration can mean a smaller injection volume. But as concentration rises, viscosity can make a biologic difficult to handle and inject. The right excipient combination depends on the molecule, the concentration and the delivery goal.
Finding that combination takes reliable experiments with difficult-to-handle materials. AEMS specializes in high-throughput viscosity screening. Those measurements let us evaluate formulations and build the models that guide further experiments.
The platform behind our work
PROPEL brings experimental design, robotic automation, and machine learning into the same process. We use each round of measurements to build a predictive model and suggest new formulations.
Define the formulation space around your molecule, delivery target, and material constraints.
Prepare formulations and measure their viscosity under defined conditions.
Use machine learning to identify promising regions, then test the suggestions in the laboratory.
Work with us today
Bring us a formulation or delivery question. We agree on a focused study, run the experiments, and help your team decide what to develop next.
Higher concentration, a smaller injection volume, or subcutaneous feasibility. Your material and product constraints define the study.
We design and run the screen, evaluate the measurements, and use PROPEL to guide further experiments within the agreed scope.
Formulation candidates with measurements, test conditions and trade-offs, plus practical recommendations for further development.
We work with excipients already used in approved drug products. Our viscosity-reduction approach does not rely on novel excipients, polymers, or enzymes. We assess ingredients against your route, concentration, and product requirements.
High-concentration proteins are difficult to handle. Automated preparation and measurement support consistent data collection across a screen, with experimental conditions kept alongside the results.
We use materials-science methods to study how ingredients work together. Testing different proportions reveals combinations that changing one ingredient at a time can miss.
We are developing predictive models from our own research data, studying formulation behavior across related molecules. The aim is to make a family-specific model more useful with each molecule studied, so future programs can start from a better-informed formulation search.
Our longer-term plan combines a free general IgG1 formulation tool, experimental validation services, and private models tailored to individual programs.
A collaboration can begin with experimental work today and help define how a predictive model would improve your pipeline. Your data stays yours, in an isolated environment.
Explore a collaborationPublished work
Scott and Alex describe how AEMS uses microscale experiments and machine learning to screen formulation mixtures.
Read the article
Chief Executive Officer & Co-Founder
Scott is a chemist and materials engineer. He completed his doctoral research at Colorado State University and postdoctoral research at Lawrence Livermore National Laboratory.
Chief Innovation Officer & Co-Founder
Alex works on experimental design and innovation at AEMS. His doctoral research at Colorado State University focused on directed evolution of metal reductases.
Tell us what you want your drug product to do: reach a higher concentration, fit a smaller injection volume, or explore subcutaneous delivery. We’ll discuss the formulation work that could help.
Coming from the PODD talk? Request a follow-up