Questions

Frequently Asked Questions

Common questions about private AI deployment, GDPR compliance, document processing, and VP Lab's interactive demos for UK businesses.

About VP Lab

What is VP Lab?
VP Lab is a collection of 12 interactive AI demos built by VantagePoint Networks, an independent IT and AI consultancy based in London. The demos showcase what private AI can do for UK businesses — document analysis, PII anonymisation, contract risk scanning, invoice extraction, email triage, RAG Q&A, and more. All demos run in your browser; private deployment runs on your own infrastructure.
Are the VP Lab demos free to use?
Yes. All VP Lab demos are free to try in your browser — no account, no credit card required. The public demos use Groq for fast inference. Private deployment on your own infrastructure is a paid service provided by VantagePoint Networks.
What is the difference between the public demo and private deployment?
The public demos send input text to Groq's API (a fast cloud inference provider) for processing. Private deployment installs open-weight language models (Mistral, Llama, or similar) on your own servers — your data never leaves your network. Private deployment is essential for regulated businesses, firms handling client confidential information, or any organisation with strict data governance requirements.
Who should use VP Lab?
VP Lab is designed for UK business owners, IT managers, legal and compliance professionals, and decision-makers at SMBs considering AI adoption. It is particularly relevant for regulated sectors: financial services, legal, healthcare, and professional services — where data privacy and compliance requirements make public AI platforms unsuitable.

Private AI Deployment

What is private AI deployment?
Private AI deployment means running a language model entirely on your own infrastructure — your servers, your private cloud, or a dedicated co-location server. The model processes your data locally, meaning inputs and outputs never travel to external AI providers. VantagePoint Networks installs, configures, and maintains private AI systems for UK businesses.
Why do regulated businesses need private AI instead of public AI tools?
Public AI tools like ChatGPT process your inputs on external servers owned by US companies. For law firms (SRA confidentiality obligations), financial services firms (FCA data standards), NHS organisations (DSPT), and any business handling personal data under UK GDPR, sending client or patient data to public AI platforms creates serious compliance risks. Private deployment eliminates these risks by keeping data within your control.
What hardware do I need for private AI?
Requirements depend on the model size and inference speed needed. For most SMB document processing workloads, a server with 32GB RAM and a modern CPU (or a modest GPU) handles Mistral 7B or similar models. VantagePoint Networks assesses your specific workload and recommends appropriate hardware — or deploys on existing servers where capacity allows.
How long does private AI deployment take?
A standard private AI deployment — model installation, configuration, integration with your document management system, and staff training — typically takes 2–5 days of VantagePoint Networks' time. Timeline depends on infrastructure complexity and integration requirements.

GDPR and Compliance

Does using AI comply with UK GDPR?
AI can be used compliantly under UK GDPR, but it requires proper governance: a clear lawful basis for processing, a Data Protection Impact Assessment (DPIA) for significant AI processing, data minimisation, and individuals' rights to explanation for automated decisions. Private AI deployment makes compliance significantly more straightforward — your data stays in your control and within UK jurisdiction.
Is the ICO concerned about businesses using AI?
The ICO supports responsible AI use and has published guidance on AI and UK GDPR. Key requirements include transparency (tell people when AI is used), fairness, accuracy, and data minimisation. The ICO has taken enforcement action against AI providers that failed to meet these standards. Using private AI on your own infrastructure addresses many ICO concerns by keeping data within your organisation's control.
Does VP Lab process any personal data?
The public VP Lab demos process whatever text you paste into them on Groq's infrastructure. Do not paste real personal data, client information, or confidential documents into the public demos. The demos are for evaluation purposes using anonymised or synthetic data. Private deployment is required for processing real personal or confidential data.
What is a DPIA and do I need one for AI?
A Data Protection Impact Assessment (DPIA) is a structured analysis required by UK GDPR before starting processing that is "likely to result in a high risk" to individuals. AI processing of personal data often triggers this requirement. VantagePoint Networks can support your DPIA process as part of a private AI deployment engagement.

Document Processing AI

How accurate is AI document analysis?
For structured extraction tasks — dates, names, monetary values, invoice numbers — accuracy is typically 95%+ on clean digital documents. For subjective analysis — risk scoring, clause interpretation, sentiment — accuracy varies and AI outputs should be treated as a first-pass review for human attention, not a final decision. Always have qualified professionals review AI outputs for legal, financial, or compliance decisions.
What document formats does VP Lab handle?
The demos handle pasted text. Private deployment handles PDF (including scanned PDFs via OCR), DOCX, TXT, and Markdown. Scanned PDFs require an OCR layer before AI processing; accuracy depends on scan quality.
Can AI replace my lawyers, accountants, or compliance team?
No. AI is a productivity tool for qualified professionals — it helps lawyers review contracts faster, helps accountants process invoices more efficiently, and helps compliance teams triage large volumes of documents. It does not replace professional judgement, legal advice, or regulatory compliance obligations. All AI outputs for professional decisions should be reviewed and validated by qualified staff.

VP Lab Demos

What AI demos are available on VP Lab?
VP Lab has 12 demos: Document Q&A (RAG), Document Analyser, PII Anonymiser, Meeting Notes to Actions, Contract Risk Scanner, Invoice Extractor, Plain English Converter, Email Triage, Document Compare, Inbound Agreement Processor, Complaints Triage, and Board Pack Digest. All are free to try in your browser.
How fast are the VP Lab demos?
The public demos use Groq's inference API, which runs Llama and Mistral models at 300–500 tokens per second — significantly faster than OpenAI's API. Most demo responses complete in 2–5 seconds for typical document lengths.
Can I use VP Lab demos with my real business documents?
The public demos are for evaluation with test or anonymised data only — do not paste real client, patient, or personal data. For processing real business documents compliantly, contact VantagePoint Networks about private deployment on your own infrastructure.

Still have questions?

VantagePoint Networks provides expert private AI consulting for UK businesses.

Try the demos free →Talk to an expert