Finance
A change in a financial market can be driven by geopolitics, supply chains, human behaviour and physical events.
Advanced Technology · Research · Intelligence
Finding signal
in complexity.
Semorion is an advanced technology company building intelligent systems for complex, high consequence environments.
We work across finance, medicine, cyber security and space, combining artificial intelligence, mathematics, software engineering, data science and domain expertise to solve problems that conventional approaches struggle to address.
Not every problem requires AI.
Our focus is broader, understand the system, identify the signal that matters, then build the right technology around it.
01 / Thesis
One company. Connected intelligence.
The hardest problems cross boundaries.
The most difficult problems rarely exist inside a single discipline.
Cyber Security
A cyber attack can emerge from millions of apparently unrelated network events.
Medicine
A deterioration in a patient's condition may be visible as a weak pattern long before conventional thresholds are crossed.
Space
A spacecraft or remote system may need to recognise an emerging failure without continuous human supervision.
These problems share something fundamental.
There is signal hidden inside complexity. Semorion exists to find it.Semorion founding principle
02 / Domains
What We Do
Four domains. One underlying challenge.
We build technology for environments where large volumes of changing information must be understood before certainty is available.
01 / FINANCE
Finance
Intelligence beyond the market feed
Financial systems are influenced by far more than price.
Semorion develops technology that can combine market data with economic indicators, company information, geopolitical events, supply chains, commodities and other signals to identify emerging changes earlier.
- Market intelligence
- Time series modelling
- Regime change detection
- Risk analysis
- Alternative data
- Continuous forecasting
- Cross domain signal analysis
- Decision support
02 / MEDICINE
Medicine
Earlier signals. Better decisions.
Important physiological changes often begin before conventional clinical thresholds are breached.
We are exploring systems designed to identify these early signals, understand changing patient states and support safer clinical decision making.
- Early warning systems
- Irregular clinical time series
- Patient deterioration detection
- Clinical model validation
- Local model calibration
- Model drift monitoring
- Remote and resource constrained healthcare
- Responsible clinical AI
Our objective is not to replace clinicians. It is to give them better information sooner.
03 / CYBER SECURITY
Cyber Security
See the attack before it becomes an incident
Modern environments generate enormous quantities of security telemetry.
The challenge is no longer simply collecting data. It is understanding how apparently unrelated events connect.
Semorion is developing systems that analyse relationships between identities, devices, infrastructure, network activity, vulnerabilities, behaviour and threat intelligence.
- Threat detection
- Attack path analysis
- Security telemetry correlation
- Identity intelligence
- Network behaviour
- Threat intelligence enrichment
- Autonomous investigation
- AI system security
- Agent governance
Our objective is to move security from isolated alerts towards connected understanding.
04 / SPACE
Space
Intelligence where connectivity cannot be assumed
Space systems operate in environments where latency, limited connectivity and physical constraints make traditional cloud dependent intelligence unsuitable.
Semorion is exploring resilient technologies capable of operating closer to the system itself.
- Autonomous monitoring
- Anomaly detection
- Telemetry intelligence
- Edge computing
- Predictive maintenance
- Distributed systems
- Sensor fusion
- Fault detection
- Mission decision support
03 / Connected Intelligence
Connected by Signal
Different disciplines. Similar structure.
Finance. Medicine. Cyber security. Space. They appear to be very different disciplines. At the level of the underlying problem, they frequently are not.
Each produces streams of observations.
Each contains noise.
Each changes over time.
Each contains weak signals that may become important later.
And each requires decisions before certainty is available.
Semorion researches the mathematics, models and systems that can operate across these boundaries.
04 / Method
Beyond Artificial Intelligence
The technology follows the problem.
AI is transforming what can be built, but it is not the answer to every problem.
Semorion combines multiple disciplines depending on the problem:
Artificial intelligence
Machine learning
Mathematics
Statistics
Time series analysis
Causal reasoning
Optimisation
Signal processing
Cyber security engineering
Distributed systems
Edge computing
Scientific computing
Software engineering
Domain expertise
The technology follows the problem, not the other way around.
05 / Labs
Semorion Labs
Research with a path to reality
Semorion Labs is where we investigate ideas that sit between research and deployable technology.
Some will become products.
Some may become scientific research.
Some will become enabling technologies used across several industries.
And some will fail.
That is part of the process.
The objective is to test difficult ideas quickly, rigorously and honestly.
Research matters when it can survive contact with reality.Semorion Labs
06 / Research
Current Research Themes
Research across changing systems.
Our current themes focus on systems that must observe change, identify weak signals and act or escalate before failure becomes obvious.
Continuous Intelligence
Observe without waiting for a request.
Most software waits for a request. Many real world systems cannot. We are researching systems capable of continuously observing changing environments, recognising meaningful change and deciding when action or human attention is required.
Regime Change
Recognise when the world has changed.
Models often perform well until the world changes. Markets change. Attack behaviour changes. Patient populations change. Sensors degrade. Operational environments shift. Semorion is researching ways of detecting these changes and adapting systems without blindly retraining them on every new observation.
Early Warning
Find the evidence before the event.
The most valuable signal is often not the event itself. It is the weak evidence that appears before the event. Our research investigates whether emerging patterns can be identified earlier across clinical, financial, cyber and physical systems.
Autonomous Systems
Control, verify and audit action.
As AI systems gain the ability to take actions, verifying what they are permitted to do becomes increasingly important. We are exploring technologies for controlling, verifying and auditing autonomous software and AI agents.
Connected Intelligence
Reason across events, relationships, time and context.
The next generation of intelligent systems will need to reason across different forms of information rather than operating on isolated datasets. Semorion is investigating architectures that connect events, relationships, time series, behaviour and external context into a coherent view of a changing system.
SEMORION
07 / Name
The Name
SEM + ORION
SEM
represents signal
+
ORION
Semorion combines two ideas.
SEM represents signal.
Orion is one of the most recognisable constellations in the night sky, historically used to understand position, direction and the wider environment.
Together they represent what Semorion is being built to do: find meaningful signals inside complex environments and use them to understand what happens next.
08 / People
The people behind Semorion
Deep expertise across the domains we work in.
Semorion brings together expertise in cyber security, medicine, mathematics and computational finance. The point is not to force those disciplines together, but to use each one where it gives us a better way to understand the system.
BH
Blanka Horvath
Mathematics · Computational Finance
Associate Professor in Mathematical and Computational Finance at the University of Oxford and researcher at the Alan Turing Institute. Her research sits at the intersection of stochastic analysis, rough path theory and machine learning for finance.
TP
Theodore Pezas
Medicine · Clinical Research
Surgeon and clinical researcher based in London with expertise in plastic and reconstructive surgery. His work brings frontline clinical judgement to the development and evaluation of medical technologies, including patient deterioration, outcomes and evidence based clinical practice.
KW
Kevin Wharram
Cyber Security · Systems · Research
Cyber security practitioner and independent researcher with more than 25 years in the field and extensive experience in SIEM and security monitoring. His work focuses on cross domain pattern recognition, security intelligence, autonomous investigation and translating research ideas into operational systems.
The team structure is intentionally cross disciplinary. Semorion is designed around problems that sit between established fields rather than inside a single technical category.
09 / Principles
Our Principles
How Semorion should work.
These principles apply across research, engineering and product development.
Signal before noise
More data does not automatically produce more understanding. We focus on identifying what actually matters.
Evidence before claims
Promising technology should survive rigorous testing. We measure systems against real baselines and real alternatives.
Problems before technology
We do not begin with AI and search for somewhere to use it. We begin with the problem.
Humans remain accountable
Intelligent systems should improve human decision making, not hide how decisions are made.
Build across boundaries
Some of the most interesting discoveries happen when methods developed in one discipline are applied carefully to another.
10 / Product Path
From Research to Products
A continuous cycle.
Semorion is designed as a laboratory and a company. Research can become intellectual property. Intellectual property can become technology. Technology can become products. And successful products can provide the resources to investigate the next generation of problems.
RESEARCH
↓
PROTOTYPE
↓
VALIDATE
↓
PRODUCT
↓
DEPLOY
↓
LEARN
↓
RESEARCH
11 / Why Semorion
Why Semorion
The problem is not more data.
The world is producing more information than organisations can meaningfully interpret.
The next challenge is not simply generating more data.
It is knowing which signals matter.
Understanding how they connect.
Recognising when something has changed.
And acting before the opportunity, failure, attack or deterioration becomes obvious.
That is the problem Semorion is being built to solve.
SEMORION
Find the signal.
Finance · Medicine · Cyber Security · Space
Research · Technology · Intelligence
semorion.com
Contact Semorion