Make every AI conversation safe, governed & accountable
CommonLayer helps enterprises detect, interpret, and respond to psychological risk inside AI conversations - before harm escalates. Built for healthcare, insurers, EAPs, governments, and AI platforms, CommonLayer acts as an intervention middleware layer between conversational systems and human care pathways.
Safer AI experiences. Real-time intervention. Human escalation when it matters most.
PLATFORM TOOLS
Everything you need to deploy AI responsibly
A comprehensive safety and governance stack that sits seamlessly between your users and any foundation model - acting as a real-time orchestration, oversight, and control layer that ensures every interaction is aligned with your policies, values, and regulatory requirements.
It continuously monitors, filters, and shapes inputs and outputs, mitigating risk, preventing harmful or non-compliant responses, and enforcing guardrails without compromising user experience.
Designed to be model-agnostic and easily integrated, it provides transparency, auditability, and adaptive control - so you can deploy AI with confidence, accountability, and trust at scale.
01
Real-Time Safety Detection
Detect crisis signals, self-harm language & emotional escalation in real time using advanced NLP before harm occurs.
02
Vulnerability Scoring
Continuously assess user risk with temporal escalation tracking and behavioural pattern recognition across sessions.
03
Governance Rule Engine
Define and enforce relational boundaries, compliance policies and configurable safety protocols for every AI interaction.
04
Audit & Compliance Reporting
Full conversation audit trails, compliance dashboards and exportable reports for regulators and clinical governance teams.
Meet our Team & Advisory Board
CommonLayer is guided by a multidisciplinary team and advisory board spanning AI technologists, mental health clinicians, researchers, behavioural scientists, product leaders, and governance experts working at the intersection of human wellbeing, mental health, and safe conversational AI.
CEO / Founder
Chris Rhyss
A military veteran, AI researcher and founder working at the intersection of mental health, conversational AI and governance, informed by PhD research into human–AI interaction and experience building award-winning health-tech.
CMO / Non-Exec Dir.
Nicki Kenyon
A globally experienced Non-Executive Director and transformation leader known for driving measurable growth, digital transformation and strategic outcomes across technology, tourism, sport, financial services and startups.
CTO (Fractional)
Lucas Sargent
Technical co-founder and fractional CTO with extensive experience building AI platforms, scaling engineering teams, and delivering secure, high-velocity technology solutions for startups and enterprises across a range of industries.
Dir. Clinical Science
Louise Metcalf
An organisational psychologist and behavioural scientist with over 30 years’ experience in psychometrics, behavioural change and AI ethics, recognised globally for applying scientific rigour to human-centred organisational systems.
UX/UI Lead (Advisor)
Oliver Weidlich
A UX strategist and researcher with more than 25 years’ experience designing human-centred digital experiences across psychology, spatial computing and emerging technologies for leading enterprises, startups and academia.
Chatbots powered by mainstream LLMs are launching at an unprecedented pace, with new apps appearing every week across app stores. What once required significant engineering can now be shipped quickly by wrapping a foundation model in a simple interface. But while capability is accelerating, responsibility isn’t keeping up - and most of these products lack the safeguards needed for real-world use.
Capability Is Scaling Faster Than Responsibility
In practice, limited thought is given to how these systems behave in edge cases, respond to vulnerability, or manage risk over time. Guardrails are often assumed to be “handled by the model,” rather than designed into the user experience.
This creates a widening gap between what these systems can do and the responsibility required to deploy them safely in human contexts.
The Safety Gap in Real-World AI Interactions
Common Layer exists to close that gap, introducing a consistent safety and governance layer between users and AI systems, embedding context awareness, clear boundaries, and real-time oversight into every interaction. Sitting at the application layer, it ensures conversations are shaped by intent and sensitivity, with the ability to guide, constrain, or escalate when needed, making safety a designed capability.
Common Layer: A System for Safe, Governed AI
WHY COMMON LAYER?
Live in minutes, not months
Common Layer deploys as a thin orchestration layer, sitting lightly across your existing systems with minimal overhead and no need for costly rearchitecting. It integrates seamlessly into your current stack, enabling coordination, intelligence, and interoperability without disrupting what already works.
01
Connect
Integrate Common Layer via API or SDK into your existing AI stack - any LLM, any platform.
02
Configure
Set governance rules, safety thresholds, escalation protocols and compliance requirements.
03
Protect
Every conversation is monitored, scored and shaped in real time - keeping users safe.
04
Report
Access audit trails, risk dashboards and compliance reports for full regulatory visibility.
Closing the Gap Between AI Capability and Responsibility
As AI becomes embedded in everyday interactions, millions of conversations are now happening in largely unregulated, opaque environments - where safety, accountability, and relational nuance are not guaranteed.
These systems can simulate empathy and authority, yet lack true responsibility, creating a growing gap between perceived care and actual safeguards. Without a dedicated layer of oversight, harmful outputs, subtle bias, or misplaced trust can scale unchecked, particularly in sensitive or high-stakes contexts.
Common Layer exists to close this gap - introducing a critical safety and governance layer that sits between users and AI, imposing relational distance, enforcing guardrails, and ensuring that every interaction is not only intelligent, but responsible, transparent, and worthy of trust.