# AI journeys overview Most teams start with a prompt and a chat window. That can be enough to test an idea, but it is usually not enough to deliver a real product experience. A useful AI experience needs structure, context, and a way to improve over time. That is what tagIt means by an **AI journey**. An AI journey is not just a single response. It is the full experience around how AI engages with a person: * why the interaction started * what the AI is trying to accomplish * how the experience is presented * what the user reveals through the conversation * how future engagement should adapt based on those signals tagIt helps teams design, embed, and improve those journeys. ## Why journeys are different from chat Chat is only the surface. The harder questions are: * What experience should this AI create? * What context should it operate within? * What tools or models should it use? * How should it behave in different scenarios? * What should it learn from the interaction? An AI journey brings those questions together into one system. Instead of thinking about AI as a generic assistant, tagIt helps teams think about: * the journey being conducted * the experience the user sees * the signals the system learns from ## The three parts of an AI journey in tagIt ### AI Orchestrator AI Orchestrator conducts the journey. It defines how the runtime should behave for a specific context (eg, website, app, campaign, service, etc) and use case. This includes: * the use case being executed * the model being used * optional tools and MCP servers * policy and runtime rules * the instructions that shape the experience This is the decision layer of the journey. ### AI Canvas AI Canvas shapes the experience. It is the optional interface layer where the journey is delivered to the user. It is how teams embed AI into products, sites, and workflows using the AI client and related experience patterns. This is the visible layer of the journey that tagIt has packaged to quickly start but you can also integrate your custom clients directly with the APIs for both AI Orchestrator and AI Presence. ### AI Presence AI Presence turns conversations into intent. It helps the system understand what a person is signaling through the interaction. Those signals can become memory, relationship insight, and engagement policy that shapes future orchestration. This is the learning layer of the journey. ## How the loop works An AI journey in tagIt works as a loop: 1. A team defines a context and use case/agent. **Think of the use case/agent as the executable runtime with specific instructions, AI model, and accessible tools.** 2. AI Orchestrator runs the experience with the right runtime configuration. 3. AI Canvas delivers that experience to the user. 4. The conversation generates meaningful signals. 5. AI Presence turns those signals into intent, memory, and insight. 6. That insight can feed back into orchestration so future interactions become more personalized. This is what makes the journey improve over time instead of resetting every turn. ## A simple example Imagine a commerce brand using tagIt to help turn an exploratory shopper into a purchaser. The brand's site **context** stays consistent across the experience. It is the top-level container for the brand, product, campaign, site, app, or agent family. It defines the broader environment and persona for the AI, such as: * you are the brand's shopping assistant * you help users discover products and make confident purchase decisions * you are helpful, clear, and conversion-focused without being pushy Within that context, the team can define different **use cases/agents** for different parts of the journey. ### 1. Homepage or landing page The user has just arrived and may not know what they want yet. The use case here might focus on: * greeting the user * helping them explore categories by connecting to catalog and category discovery tools * understanding what they are shopping for * narrowing broad interest into product intent At this stage, the goal is not to force a purchase. It is to help the shopper move from exploration to clearer intent. ### 2. Product listing or category page Now the user is comparing options. The use case can shift to: * recommending relevant products by connecting to the MCP-powered product catalog * highlighting differences between options by using the product attribute tool * answering lightweight fit questions * helping the user narrow the shortlist with inventory and merchandising signals The same context is still in place, but the runtime behavior changes because the user is in a different part of the journey. ### 3. Product detail page Now the user is much closer to a purchase decision. The use case can become more specific: * provide deeper product information from product detail and specification tools * compare this product to alternatives using comparison and attribute MCP tools * answer objections * clarify features, pricing, or compatibility by retrieving policy, pricing, and fit data * encourage a confident next step This is where AI Presence becomes especially useful. If the user keeps asking for comparisons, more detail, or reassurance, those conversation signals can reveal what kind of engagement will help move them forward. ### 4. Cart or checkout stage At this point, the goal is to reduce friction. The use case can focus on: * answering last-minute questions * clarifying shipping, pricing, or return policies through checkout, shipping, and policy tools * helping resolve uncertainty before purchase The journey is still continuous, but the AI is now working with a different objective than it had on the homepage. ### 5. Post-purchase After the order is placed, the journey does not have to end. The use case can shift again to: * delivery and order updates using order and shipment tools * setup or onboarding guidance by retrieving the right support or help content * support and follow-up * cross-sell or repeat-purchase opportunities when appropriate This is a good example of why tagIt thinks in journeys instead of isolated chats. The same customer can move through multiple moments, each with a different use case, while still staying inside one broader relationship context. ### Why this matters Without this structure, the AI is just reacting turn by turn. With tagIt: * the **context** gives the AI a stable role and environment * the **use case** tells the AI what it should accomplish in the current moment * the **MCP tools** available to that use case determine what actions and retrieval paths it can use in that stage of the journey * **Presence** helps the system learn from what the shopper is signaling * future engagement can become more relevant as the journey unfolds That is the difference between a generic assistant and an AI journey designed to guide someone toward an outcome. ## Why this matters Teams do not just want AI to answer questions. They want AI to: * guide users * adapt to signals * stay aligned to the experience they are designing * produce insight that improves future engagement tagIt is built for that full loop. ## In one sentence An AI journey is the full loop of orchestration, experience, and presence that turns one interaction into a smarter next interaction. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://tagit.gitbook.io/tagit/getting-started/ai-journeys-overview.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.