Where we started
Quantitative Pharmacology and Systems Biology
PK/PD modelling, QSP and model-informed drug development: quantitative understanding of how drugs and biological systems behave.
Explore pharmacologyQuantitative research and AI engineering
Deep research for problems that resist simple answers.
Xenalys is a quantitative research company. We began by modelling the complexity of biological/pharmacolgical systems, and we bring the same abstraction and rigour to problems in finance and business, using AI to move faster without losing traceability. Our goal is not to develop sophisticated computational or mathematical models or AI frame works for their own sake, but to solve complex problems by identifying and applying the most appropriate approaches.
Many interacting factors
Abstraction
A traceable model
Our technology foundation
Advanced Computational Methods. Intelligent Systems.
Our business areas are supported by a shared technology foundation combining scientific methodologies, quantitative approaches and artificial intelligence.
Artificial Intelligence
Machine Learning
Statistical Modelling
Computational Methods
Intelligent Automation
Our origin
Complex systems are among the hardest things to model. Doing it well takes imagination and abstraction, and the discipline to keep every model traceable. That is the research culture we bring to every domain.
Abstraction
Distilling a huge number of interacting factors into a mathematical formalism that stays tractable.
Traceability
Creative models only earn trust when assumptions, methods and limits are visible and reproducible.
AI with judgement
AI accelerates the work. Research experience decides what is correct.
Our domains
One research culture, four domains.
Quantitative Finance
Alpha and factor research, portfolio optimisation, risk and credit analytics, and agentic research systems for institutions.
Explore financeBusiness and Decision Intelligence
Market sizing, feasibility, profitability and operational analysis, turned into repeatable decision workflows.
Explore business intelligenceAI/ML and Agentic Engineering
Machine learning, retrieval systems and AI agents with tool use, multi-step workflows and rigorous evaluation.
Explore AI engineeringHow we work
Two ways we work. The same rigour in both.
DEEP RESEARCH
For novel, ambiguous, high-stakes problems.
- Problem framing and abstraction
- Bespoke mathematical and statistical models
- Creative solutions, tested before they are trusted
- Researcher-led from question to result
ROUTINE SERVICES
For well-defined, repeatable analysis.
- Agentic AI frameworks run the workflow
- Minimal human assistance, so faster and more cost-effective
- Research-grade checks on every output
- Repeatable and reproducible by design
Our method
From a hard question to a result you can check.
Step 1
Frame the problem
Define the question, the decision it supports and what a good answer looks like.
Step 2
Abstract and model
Reduce the complexity to the factors that matter and express them in a formal model.
Step 3
Validate independently
Test assumptions, diagnostics and robustness, and state the limits plainly.
Step 4
Deliver and enable
Hand over reproducible outputs and, where useful, a repeatable AI-enabled workflow.
Our commitment
Scientific Rigour. Analytical Accuracy. Technological Reliability.
We believe that scientific and technological advancement must be accompanied by methodological discipline, transparency and measurable reliability.
Scientific Rigour
Applying established scientific and analytical methodologies.
Accuracy
Prioritising reliable results, validation and evidence-based reasoning.
Transparency
Maintaining clarity in assumptions, methods, limitations and outcomes.
Reproducibility
Building traceable and reproducible analytical and computational workflows.
Efficiency
Using intelligent systems to improve productivity and operational performance.
Cost-effectiveness
Making scientific and technological solutions more accessible and economically efficient.
Contact
Bring us a problem that is hard to frame.
Tell us the question you are trying to understand or decide. We will propose a research-led approach.
Start a conversation