DeepCyte, a techbio company developing AI-driven toxicology tools for drug development, has launched with $1.5 million in seed funding. The company is introducing two core solutions designed to help biopharma teams detect, predict, and explain drug toxicity in human cells at single-cell resolution.

Drug-induced toxicity remains one of the leading causes of clinical trial failure and post-market drug withdrawal, contributing to significant cost and attrition across the pharmaceutical industry. Traditional approaches such as animal models, high-throughput screening, and bulk assays often fail to accurately predict human biological responses and are limited in their ability to capture cell-to-cell variability. At the same time, regulators including the FDA and EMA are increasingly encouraging the adoption of more human-relevant, mechanism-aware approaches to safety assessment, supporting demand for more predictive technologies in preclinical development.

“DeepCyte’s mission is to reveal and prevent toxicity in every cell, at scale, before drugs reach patients,” said Theodore Alexandrov, Ph.D., CEO and co-founder. “By combining advances in AI and single-cell biology, we predict not only whether a drug is toxic, but also why.”

The company’s first platform, MetaCore, is a high-throughput single-cell metabolomics system built on laser-based sampling and mass spectrometry technology. It generates detailed molecular profiles that capture cellular heterogeneity often missed by bulk analysis methods. DeepCyte says MetaCore is designed to produce large-scale, AI-ready datasets with minimal sample preparation and at a cost intended to support broader adoption in drug discovery and translational research workflows.

Building on this dataset layer, DeeImmuno is DeepCyte’s first AI application, using machine learning models trained on proprietary single-cell metabolomics atlases generated through MetaCore. The system is designed to classify toxicity, identify biomarkers, and infer underlying molecular mechanisms. In internal evaluation across 100 held-out compounds, DeeImmuno reported 94% accuracy in predicting 17 detailed toxicity mechanisms, offering what the company describes as high-resolution mechanistic insight not typically achievable with conventional toxicology approaches.

DeepCyte’s leadership team brings together expertise in pharmacology, artificial intelligence, and life sciences commercialisation. Chief executive Theodore Alexandrov previously developed METASPACE, a cloud-based software platform used by researchers globally, and co-founded SCiLS GmbH, which was later acquired by Bruker (NASDAQ: BRKR). The seed round was supported by medtech executive and investor Carl J. G. Evertsz, who will serve as board chair.

Speaking on the launch, Alexandrov said: “DeepCyte’s mission is to reveal and prevent toxicity in every cell, at scale, before drugs reach patients.”

The company positions its approach as part of a broader shift toward single-cell biology and AI-enabled predictive toxicology, aiming to improve early-stage decision-making in drug development and reduce late-stage failure rates.