Mathematical modelling
Operator-based and structure-aware representations of complex molecular behaviour.
Advanced computational technologies
Engineering the future of computational molecular science.
Spacefield develops mechanism-informed technologies that combine mathematical modelling, scientific computing, molecular intelligence and artificial intelligence to accelerate research and support experimental discovery.
Mathematical modellingScientific computingMolecular intelligence
Computational technologies
A rigorous foundation for molecular intelligence.
Our work connects mathematics, software engineering and molecular representation to create research systems designed for disciplined, traceable scientific use.
Scientific computing
Engineered computational workflows for reproducible analysis and scalable prediction.
Molecular intelligence
Integrated descriptors, models and ranking systems for evidence-led prioritisation.
Mechanism-informed prediction
Prediction methods informed by chemical structure, physical relationships and validated computation.
Technology stack
One connected computational ecosystem.
Spacefield is the company. Computational science is the capability. The SFT Molecular Platform is the flagship application.
Company
Spacefield
Deep computational technology.
→
Foundation
SFT-Core & descriptor engines
Computational representations, preprocessing and scientific traceability.
→
Flagship application
SFT Molecular Platform
SFT-pKaSFT-logPSFT-SolubilitySFT-BBBSFT-ADMETSFT-Binding
Scientific research outputs
Far beyond a single prediction.
The platform is designed to generate structured scientific artefacts for analysis, traceability, reproducibility and integration—not merely CSV, JSON and PDF files.
Prediction outputs
- CSV datasets
- JSON records
- PDF reports
Computational outputs
- Prediction & runtime logs
- Validation & benchmark logs
- Audit and error records
- Execution summaries
Scientific outputs
- Candidate ranking
- Descriptor summaries
- Confidence analysis
- Comparative benchmarks
Reproducibility outputs
- Frozen manifests
- Integrity hashes
- Dataset fingerprints
- Configuration snapshots
Integration outputs
- REST API responses
- SDK exports
- Batch & CLI reports
- Workflow artefacts
Scientific methodology
From mathematics to validated discovery.
A transparent workflow keeps computation connected to scientific reasoning and experimental evidence.
Mathematics
→
Computation
→
Prediction
→
Validation
→
Discovery
Scientific position: computational predictions are used to prioritise experiments, focus resources and generate testable hypotheses. They do not replace synthesis, biological assays, safety studies or independent experimental validation.
Collaboration & experimental validation
Turning computational candidates into experimental evidence.
We welcome conversations with CROs, universities, academic laboratories, pharmaceutical R&D teams and research institutes capable of supporting synthesis, assay development, biophysical testing, ADMET studies and preclinical validation.
Discuss a collaboration