Frequently Asked Questions

AI Prompt Architect is an enterprise-grade SaaS platform for professional prompt engineering. It helps developers, teams, and businesses design, test, and deploy AI prompts with a structured, repeatable workflow — turning ad-hoc prompting into a disciplined engineering process.

It's designed for software developers, AI engineers, product teams, legal professionals, and any business that uses large language models. Whether you're building chatbots, automating document generation, or integrating AI into your product — this platform gives you the tools to get consistent, high-quality results.

No. The platform features a guided wizard that walks you through prompt construction step by step. However, developers will benefit from advanced features like system context management, variable injection, and API integration.

Currently we support Google Gemini models including Gemini 2.5 Pro and Gemini 2.5 Flash. We use Google's latest models to deliver the highest quality output. Additional model providers may be added in the future.

Absolutely. Vendor lock-in goes against our core architecture philosophy. You can export your finalized prompts and system contexts into standard Markdown or JSON formats for use in any CI/CD pipeline, IDE, or external AI agent.

We take data security seriously. Your prompts are encrypted using AES-256-GCM — the same standard used by banks and government agencies. If you use our BYOK (Bring Your Own Key) mode, your API key is encrypted in your browser before any storage using a zero-knowledge architecture — meaning even our own team can't see your key. During API calls, keys exist in server memory for less than 2 seconds and are never logged. All traffic is encrypted via TLS 1.3. We do not use your prompts to train our own models.

Zero-knowledge means we literally cannot see your data, even if we wanted to. When you save your API key, it's encrypted in your browser using AES-256-GCM with a key derived from your unique user ID. Only your authenticated browser session can decrypt it. Our servers, administrators, and database operators only ever see an encrypted blob — never the plaintext key.

No. This is by design. Our admin panel displays "[ENCRYPTED]" for all user API keys. No staff member — including the founder — can view, access, or recover your plaintext key. This is enforced at the architecture level, not just by policy.

When you make an AI generation request, your key is decrypted in your browser, transmitted over TLS 1.3 to our secure Cloud Function, used to call the AI provider, and then immediately released from memory. The entire process takes less than 2 seconds. The key is never written to disk, never logged, and never backed up.

No. We do not use your prompts, inputs, or outputs to train any AI models — ours or anyone else's. Your intellectual property remains yours. This is explicitly stated in our Terms of Service and Privacy Policy.

Yes. We comply with GDPR, UK GDPR, and the Australian Privacy Act 1988. We act as a Data Processor on your behalf. You retain full control over your data, including the right to access, correct, and delete it. Our Data Processing Agreement is available in our footer.

In the unlikely event of a security incident affecting BYOK keys, we will: (1) Notify all affected users within 72 hours as required by GDPR, (2) Recommend immediate API key rotation, (3) Report to the ICO and relevant data protection authorities. Full details are in our Privacy Policy Section 7.

Yes! Our Free Architect tier includes complimentary credits so you can explore the platform and build your first prompts at no cost. No credit card required.

Credits are our internal currency for AI operations. Different actions (analysis, generation, refinement) consume different amounts of credits. Once you've used your credits, you can purchase additional credit packs or upgrade to a higher tier for more credits. Credit balances are visible in your dashboard at all times.

It depends on your plan. Our Pro Architect tier includes a specific number of analysis and generation credits directly through our platform (no keys required). If you choose the API Tier, you will need to provide your own Google Gemini API key but you will pay zero markup on the usage.

We support multiple currencies including USD, GBP, EUR, AUD, NZD, and CAD. Pricing is displayed in your local currency where available, and all payments are processed securely through Stripe.

Credits are non-refundable once purchased as they are immediate-access virtual goods. However, if you experience a technical issue that prevents you from using your credits, please contact our support team and we'll work with you to resolve it.

Credits are valid for eighteen (18) months from the date of purchase. We also reserve the right to expire credits if your account is inactive for twelve (12) consecutive months. We will notify you at least 30 days before any expiry occurs.

BYOK (Bring Your Own Key) is our API Tier where you provide your own Google Gemini API key. Instead of using our credits, your AI requests go directly through your key — meaning you pay Google directly at their standard rates with zero markup from us. You get unlimited generations while retaining full control over your API usage and billing.

Visit Google AI Studio (aistudio.google.com/app/apikey) and sign in with your Google account. You can create an API key in seconds — it's free to get started, and Google offers a generous free tier. Once you have your key, paste it into the API Keys section of your Settings panel.

Yes. Your key is encrypted in your browser using AES-256-GCM before it ever leaves your device. The encryption key is derived from your unique Firebase user ID using PBKDF2 with 100,000 iterations. The encrypted blob is stored in our database — but we cannot decrypt it. Only your authenticated browser session can.

No. There is zero markup on API usage when using BYOK. You pay Google directly at their published rates. The only cost from us is the API Tier subscription fee (if applicable), which gives you access to the platform features.

If you suspect your key has been compromised: (1) Immediately revoke it in Google AI Studio and generate a new one, (2) Update your key in the Settings panel. Since we use zero-knowledge encryption, a breach of our database would NOT expose your plaintext key. However, you are responsible for key security on your own devices.

AI Prompt Architect provides a structured engineering workflow — not just a chat box. Features include: multi-section prompt composition, system context management, variable injection, output format control, legal document generation, compliance checking, prompt versioning, and team collaboration. It turns prompting from trial-and-error into a repeatable engineering discipline.

Yes. We have a dedicated Legal Assistant module that generates jurisdiction-specific legal documents (Terms of Service, Privacy Policy, EULA, etc.) using our curated legal knowledge base of 1,000+ compliance requirements across US, UK, EU, AU, NZ, and CA jurisdictions.

The platform is currently web-based. We're exploring API access for enterprise customers who want to integrate prompt execution into their CI/CD pipelines. Contact us if you're interested in early access.

The web application is responsive and works on tablets. For the best experience, we recommend using it on a desktop or laptop. Native mobile apps are on our roadmap.

Go to Settings → General → Danger Zone → Delete Account. You'll be asked to type "DELETE" to confirm. This permanently removes all your data, credits, and history. This action is irreversible.

You can email us directly at support@aipromptarchitect.co.uk or use the Report/Feedback section in your Settings panel. We aim to respond within 24 hours on business days.

Your data is stored on Google Cloud Platform (Firebase) infrastructure, primarily in US-Central and Europe-West regions. All data is encrypted at rest and in transit. For BYOK keys specifically, we use client-side AES-256 encryption so the stored data is doubly encrypted.

Yes! We offer an affiliate program where you can earn commissions by referring new users. Visit our Affiliate page for details on commission rates and how to get started.

Development AIs (like Cursor, Claude, or Devin) lose context when given vague initial instructions — a phenomenon known as "context collapse." Without strict constraints, the AI guesses your architecture and drifts off-track as the session lengthens. To prevent this, developers use a Master Prompt — a highly structured document that explicitly locks the AI into a specific tech stack, file structure, and set of coding standards.

An AI Master Prompt is a comprehensive set of enterprise-grade instructions fed to an AI development assistant before coding begins. Unlike a simple text request, a Master Prompt defines strict rules for the AI's role, the required frontend/backend frameworks, exact folder structures, and execution steps — ensuring the AI stays on track and does not hallucinate incompatible libraries.

Writing a perfect system prompt manually is time-consuming and error-prone. The most efficient method is to use a Master Prompt Generator, like AI Prompt Architect. You simply describe your desired application in plain English, and the engine automatically generates a structured, granular Master Prompt that you can paste directly into Cursor, Claude, or any AI coding assistant to keep it strictly constrained to your project's rules.

Yes. When you describe your application to AI Prompt Architect, you can specify any language, framework, or database (such as React, Next.js, Python, Django, or Firebase). The engine will generate a Master Prompt with architectural rules and file structures perfectly tailored to that specific tech stack, ensuring your AI developer writes compliant, idiomatic code.