Fred in three steps.
Install. Train on your content. Go live, with guardrails already on.
From setup to launch.
Step 1
Install Fred
Add Fred to your site. No developer, no code. The setup wizard walks you through it.
Step 2
Train on your content
Point Fred at your pages and documents. It learns your services, your hours, and the way you do business.
Step 3
Go live, guardrails on
Test Fred with your content, selected guardrails, and integrations. Check contact delivery and escalation, then launch when the setup meets your requirements.
Set it up once. Fred works every shift after.
Talk to Fred.
What Fred does behind the scenes.
Content retrieval, guardrail settings, and integrations support Fred’s conversations. Here is how those parts fit together and what to configure for your business.
Reads the room
Fred figures out what a visitor actually wants, not just the words they typed, so it answers the real question.
How Fred understands visitors →Knows your business
Fred learns from your website, your documents, and your sources, so answers sound like you and stay accurate.
How Fred learns your content →Built to stay in scope
Industry guardrails help define Fred’s conversation boundaries. Configure and test the rules for your business, including requests that should go to your team.
How Fred stays in line →Uses the right AI
Fred picks the best model for the job from OpenAI, Anthropic, Google, or xAI, on your own keys, with no markup.
How Fred picks the model →Hands off your leads
Connect captured inquiries to your CRM and test delivery, so your team has the information it needs for follow-up.
How Fred connects to your CRM →Knows who to call first
Use conversation summaries and lead signals to help your team prioritize follow-up.
See lead scoring in action →Why Fred is reliable enough for your homepage.
Fred brings together configured rules, answers grounded in your content, and lead-capture workflows. Review the settings and test the conversations your business expects before launch.
Configured guardrail checks
Configured checks evaluate messages and responses outside the model and can block detected matches. Test their coverage with your content, provider, and channels.
Answers from your content, not the open internet
Fred uses your pages, documents, and approved sources to ground answers. It is designed to acknowledge gaps and support follow-up with your team. Review responses for unsupported statements.
Security your legal team can review
Review data handling, consent settings, and security requirements for the deployment you choose. Contact us with the controls your reviewer needs to verify. trust page
Economics that do not punish you for scaling
Bring your own AI provider keys and pay your provider directly at its published rates. Fred charges a platform fee; your usage and connected services determine additional costs.
Go deeper on the parts that matter to you.
Guardrails
- Rules evaluated outside the model
- Industry guardrail packs and your custom rules
- How the discipline works →
Your Content
- Fred answers like you, because it learned from you
- Your site, your documents, your catalog
- How training works →
Buying Signals
- Intent, sentiment, and urgency on every conversation
- Your team knows who to call first
- How Fred reads the room →
AI Models
- Your keys, your provider, no markup
- OpenAI, Claude, Gemini, or Grok underneath
- How BYOAPI works →
See Fred work on your site.
Choose a plan, configure Fred for your business, and test before launch. Or ask Fred a question in the corner of this page.