AI-driven hit discovery
We identify promising bioactive compounds using physics-based predictive modelling and tailored generative AI tuned to agrochemical endpoints.
Services
AI-Driven Experimental Pipelines for Agriculture R&D
At Upgrow we deliver AI-enabled discovery, modelling, and validation services that shorten the path from concept to field-ready agrochemical and bio-based modulators.
Cheminformatics & molecular modeling
We perform in silico screening across public and proprietary datasets, including natural products, to prioritise safe and effective formulations.
Experimental assessment & testing
We validate growth modulators across plant-based systems and run in vivo / in vitro assays covering biochemical, physiological, and genetic properties.
Who we are
Multidisciplinary scientists advancing bioactive innovation
We are Upgrow, a University of Warsaw spin-off where researchers, engineers, and innovators combine biology and information technology.
Our mission is to empower biotech research through cutting-edge AI, cheminformatics, and plant-based experimental systems. We envision a future where data-driven generative approaches streamline the delivery of safe, effective, and sustainable solutions.
By uniting advanced computation with hands-on lab experience we accelerate the discovery and assessment of agrochemical compounds. We specialize in three core technologies:
- Data-driven prediction and generative design using chemical foundation models.
- Assessment of growth modulators, additives, and natural preparations.
- Exploration of microbial interactions across plant and environmental systems.
Projects & publications
Innovating with agrochemical data
Our research team publishes rigorously curated datasets and studies that help the community evaluate molecular properties and assess safety profiles with confidence.
ApisTox
ApisTox provides a dedicated benchmark for classifying small molecule toxicity on honey bees, supporting safer agrochemical development and regulatory reviews.
MolPILE
MolPILE is a large-scale, diverse dataset of 222+ million compounds, demonstrating how retrained foundation models improve molecular representation learning for agrochemistry.
Targeted ML models
We evaluate AI and ML models focused on agrochemical challenges, including predicting pesticide toxicity to honey bees. We highlight the need for targeted model development geared toward the agrochemical domain.
Generative tools for agrochemistry
Foundation models tailored to natural products and bio-based modulators
Chemical foundation models have transformed biology and medicine. We extend these capabilities to industrial chemistry and agrochemistry, including natural products and bio-based growth modulators.
By learning from vast collections of chemical, biological and environmental data, our models surface non-obvious relationships, predict properties and toxicity, and accelerate the design of safer, more effective products.
New candidate compounds
Searching for next-generation plant growth regulators
We apply our ML/AI toolset to discover candidates with strong agronomic potential. Several molecules are under evaluation, including a derivative of natural substance TSGM.
Get in touch
Let's talk about growing your business
Drop us a line or scan the QR code to contact our experts.
Company: Upgrow sp. z o.o.
Email: contact@upgrow-ag.com
Address: Miecznikowa 1/309c, 02-096 Warsaw, Poland
Meet us: We are at ABIM 2025 - The Annual Biocontrol Industry Meeting in Basel,
Switzerland.
