This policy explains how Apex Artificial Intelligence Limited builds and operates artificial intelligence responsibly, the technologies we rely on, and the safeguards we apply.
Who we are
Apex Artificial Intelligence Limited (“Apex AI”, “we”, “us”), a company registered in England and Wales (company number 17194741), registered office 167-169 Great Portland Street, 5th Floor, London, W1W 5PF. Contact: ash@apexaiuk.com.
The AI we use
We build on leading, vetted third-party AI providers rather than training our own foundation models. Our systems are assembled from:
- Anthropic (Claude) for natural-language understanding and generation, used to power conversation, drafting and reasoning in our agents.
- Vapi for real-time voice infrastructure, which carries the AI receptionist's phone conversations (speech recognition, turn-taking and speech synthesis).
- HighLevel (GoHighLevel) for CRM, scheduling, messaging and workflow automation, where leads, bookings and follow-ups are stored and actioned.
Each provider is engaged under its own terms and data-processing agreement. We configure these tools around each client's services, tone and rules; we do not deploy generic, unsupervised bots.
Human oversight
AI handles routine, repetitive tasks such as answering calls, booking appointments and capturing enquiries. We design clear escalation so that anything urgent, complex, clinical or sensitive is routed to a human, and a person remains accountable for the service. AI is never left to make high-stakes decisions on its own.
Accuracy and limitations
AI systems can make mistakes. We mitigate this by scoping each agent to defined tasks, testing before launch, and monitoring and tuning over time. Our agents do not provide medical, legal or financial advice, and they hand off to a human when a request falls outside their scope.
Transparency
We do not set out to deceive callers. If a caller asks whether they are speaking to a machine, the agent answers honestly that it is an AI assistant acting for the business.
Data, privacy and model training
Personal data handled by our AI, including call audio, transcripts and contact details, is processed in line with our Privacy Policy and Data Protection statement. Where the option exists, we ensure client and end-customer data is not used to train our providers' general models, and we bind sub-processors to appropriate confidentiality and security obligations.
Security
Data in transit and at rest is encrypted, access is restricted on a need-to-know basis, and we review our providers' security posture as part of selecting them.
Acceptable use
We will not build or operate AI for unlawful, deceptive, discriminatory or harmful purposes, and we reserve the right to decline or stop work that conflicts with this policy.
Bias and fairness
We are alert to the risk of biased or unfair outcomes from AI and take reasonable steps to detect and reduce them in the way we configure and monitor our systems.
Accountability
Questions, concerns or reports about our use of AI can be sent to ash@apexaiuk.com. We review this policy as our tools and the law evolve.
This document is provided as a general template and is not legal advice. Please have it reviewed against your specific circumstances before you rely on it.