The AI Engine for
Any Pathology Question
CipherX turns pathology into structured, explainable intelligence, powering diagnostics, research, and drug development.
growing daily.
The AI core that turns tissue
into knowledge.
CipherX is the technology platform behind Deciphex AI - an engine that supports many workflows, products and partnerships without being limited to any single use case.
Whole slide pathology
CipherX processes any whole-slide image at full resolution and produces outputs the pathologist can examine.
See the morphology
Broad AI representations learn numerical representations of pathology across tissue, species and disease states.
Speak pathology
Those representations are translated into human-readable signatures.
Explainable intelligence
CipherX turns tissue patterns into outputs that can
be reused, adapted and inspected.
CipherX is already deployed in our
clinical and research workflows.
Those deployments demonstrate what the engine can enable today. One engine, proven twice, in two of the hardest markets in pathology: live clinical diagnosis and regulated drug safety.
AI-powered digital pathology for hospital laboratories and NHS Trusts, supporting the 28-day Faster Diagnosis Standard. Every case signed out by a pathologist.
in production
in production
Why CipherX for diagnostics
CipherX runs an independent read on every case and points the pathologist to the tissue it flagged. The adenocarcinoma and melanoma signatures in production return 99.85% and 98.76% negative predictive value. Your pathologists sign out faster, and nothing leaves the lab unchecked.
GLP-compliant digital pathology used by 9 of the top 10 global pharma. Pre-trained lesion-screening models cut pathologist review time.
Why CipherX for research
With CipherX, you can now read out translational and clinical research studies in days instead of weeks, with every finding decomposing to named tissue structures your regulators can audit. Science that compounds across every study you run.
Built to understand pathology,
not just classify images.
CipherX is designed around breadth, interpretability and adaptability - the qualities required for pathology AI to move beyond narrow point solutions.
Comprehensive Coverage
Across tissues, species, modalities, and workflows. Models inside CipherX have been trained and validated across more pathology contexts than any comparable system.
Explainable by Design
Interpretable semantic features at every level. Pathologists understand why a prediction was made - not just what it says.
Adaptive Intelligence
Reuse, don't retrain. The CipherX engine adapts to new applications by recombining existing semantic building blocks rather than starting from zero.
Reusable Infrastructure
One engine supports multiple applications across research, clinical trials, and diagnostics.
Pathology companies already run on CipherX.
CipherX is validated where it matters: in production, across our own clinical and research workflows. Deciphex is the first and largest user of CipherX - the same engine our partners use is the one we rely on every day, signed out by expert pathologists.
AI built where pathology happens
CipherX is grounded in Deciphex's operational experience across diagnostic and research pathology. It is not developed in isolation from the workflows it is intended to transform.
From question to application
Whether the goal is a biomarker, a companion diagnostic, a research workflow or a future product, CipherX provides a common engine for converting tissue patterns into usable intelligence.
Why CipherX,
and why now?
Every result it returns decomposes back into named structures a pathologist can read and audit. Regulated workflows need that anchor into the tissue.
Our semantic layer gives tissue features and patterns names that pathologists can recognise and work with. These become the basis for our tools, and allow us to trace their conclusions back to the features behind them. As foundation models improve, we can adopt them while preserving the knowledge pathologists have already helped define.
That has been the barrier to adoption in clinical reporting and in the pharma workflows where every claim has to be defensible. CipherX is the engine we have built to overcome this barrier, and it is where we believe the next generation of clinical and research pathology will be built.
We don't just build models. We own where pathology AI is trained, validated, and deployed, across hundreds of thousands of live cases a year.
CipherX is built to answer any pathology question.
Partner with us. See how it can answer yours.