Products

AI should know the information that matters to you.

A general AI knows a lot about the world. That does not mean it knows your material, your terminology, your domain, or which sources you actually trust.

HALbot is about closing that gap.

HALbot

Your information. Useful AI.

HALbot is an approach to AI built around a specific body of information. That information might be documents, reference material, procedures, technical knowledge, teaching material or another bounded knowledge domain.

Instead of relying only on what a general-purpose model happens to know, HALbot retrieves from the selected information and uses AI to reason with it in context.

The point is not “a chatbot with a clever name.” The point is AI that can work with information you actually care about.

your documentsdomain knowledgetrusted sourcescontextual answers
DanceBot

HALbot for ballroom dance

DanceBot is the first domain-specific expression of the idea: AI working with a carefully organized body of ballroom-dance knowledge rather than guessing from the open internet.

Its job is not to replace a dance teacher. It is to make useful technical knowledge easier to find, question and explore.

DanceBot is a specialized HALbot. HALbot is not a component of HALWorld. They are separate parts of the wider HALbot Project.

Explore StepsToDance →
What could a HALbot know?

The pattern is deliberately reusable.

A company's knowledge

Policies, product information, processes, support material, internal references and operational knowledge.

A specialist domain

Curated technical, professional, educational or subject-matter information where provenance and terminology matter.

A person's information

A bounded collection of material somebody wants to search, understand, compare or work with using AI.

HALbot asks: what changes when an AI can work with the information that matters to you?