What admins can see
Message count, timestamps, lead status, sentiment/category, and action-needed indicators.
Private-by-design AI infrastructure
Private conversations should stay private. Even we should not be able to read your private messages.
Liminal is built on a simple rule: private conversations should not become someone else’s dataset, dashboard, or evidence pile.
Liminal gives site operators only what they need: message count, timestamps, lead status, sentiment/category, and whether action is needed.
Message content stays encrypted and privately owned by the client/user. Liminal operators cannot access private conversation bodies by default.
Your conversations are not our product. Liminal does not sell, mine, or expose client conversations.
Private by Design
Privacy is the core trust layer, not an add-on.
Site owners should not be able to read visitor message bodies. Admin access is limited to operational metadata needed for follow-up and quality control.
Even we should not be able to read your private messages.
Your conversations are not our product. Liminal does not sell, mine, or expose client conversations.
What admins cannot see
Raw conversation bodies and private user message text in dashboard views.
How data is handled
Minimum operational metadata is retained, with encrypted/client-side protected message handling where possible.
Read the full model: Privacy Architecture
Who it is for
Built for service businesses that rely on leads
Liminal is designed for clinics, consultants, agencies, local service businesses, and founder-led companies whose websites need to do more than look good.
Clinics and wellness businesses
Make intake clearer and reduce drop-off before staff follow-up.
Agencies and consultants
Qualify fit earlier and route prospects toward the right offer.
Local service providers
Turn local traffic into structured inquiries instead of missed chances.
Founder-led companies
Keep your voice while creating a repeatable conversion system.
What it does
From website traffic to qualified conversations
Capture
Turn passive visitors into active conversations.
Understand
Learn what users want based on behavior, intent, and context.
Convert
Guide visitors toward booking, buying, contacting, and becoming long-term clients.
Why Liminal
This is not another chat box
ChatGPT and Claude answer prompts. Liminal runs inside your business experience.
| Generic AI tools | Liminal |
|---|---|
| Answer questions | Guides visitors toward outcomes |
| Require users to know what to ask | Adapts to business context |
| Sit outside your business workflow | Captures leads and behavior |
| Do not own the customer journey | Supports private deployment and rollback/version control |
Liminal enforces metadata-only admin visibility by default so private conversation content stays private.
Liminal does not replace your team. It prepares every conversation so your team can perform at their best.
Proof / Use Case
Real use case: Privacy-first lead operations
A growth-stage team needed more than a static site. Liminal turned the experience into a private-by-design lead capture and follow-up system with behavioral signals, structured intake, and action-ready routing.
Before
No reliable lead capture or journey visibility.
After
A live funnel with tracked submissions, drop-off insight, follow-up paths, and admin visibility.
Result
A system built to measure where visitors engage, where they fall off, and how to improve conversion over time.
How It Works
How implementation works
1. Map the business
We identify the offer, audience, objections, and conversion path.
2. Build the system
Liminal creates the frontend experience, lead flow, tracking, and business-specific persona.
3. Improve from real behavior
The system tracks engagement and drop-off points so the site can keep improving.
Trust
Built for trust first
Liminal enforces metadata-only admin visibility by default so private conversation content stays private.
Your conversations are not our product. Liminal does not sell, mine, or expose client conversations.
Even we should not be able to read your private messages.
See exactly how this works in our Privacy Architecture page.