Merge pull request #6 from ibm-client-engineering/eashan-updates

deleted overview
ibm-client-engineering · Oct 15, 2024 · 97bd039 · 97bd039
2 parents c5293a6 + 897062f
commit 97bd039
Show file tree

Hide file tree

Showing 6 changed files with 21 additions and 19 deletions.
diff --git a/assets/screenshots/disam_actions_step.png b/assets/screenshots/disam_actions_step.png
diff --git a/docs/01-Business Value/01-Overview.mdx b/docs/01-Business Value/01-Overview.mdx
diff --git a/docs/01-Business Value/02-Use Cases/01-IT Helpdesk.mdx b/docs/01-Business Value/02-Use Cases/01-IT Helpdesk.mdx
@@ -18,17 +18,17 @@ The IT Helpdesk with AI-Powered Virtual Assistant is a cutting-edge solution tha
 - **Generative AI-Powered**: watsonx assistant is powered by Generative AI, enabling it to understand and respond to natural language queries.
 - **Watsonx.ai, Watson Discovery, and Watson Assistant**: The virtual assistant leverages the power of Watsonx.ai, Watson Discovery, and Watson Assistant to provide accurate and reliable resolutions.
 
-### Disambiguation Methods
+### [Disambiguation Methods](/Create/Disambiguation%20Methods)
 Leveraged LLMs to create various topics in the assistants, enabling actions to be accurately mapped to the correct options. This improved the ability to clarify user questions and enhanced overall disambiguation within the assistant.
 
-### Support Ticket Management
+### [Support Ticket Management](/Create/Support%20Ticket%20Management/solution-prepare-optional-service-now)
 - **ServiceNow Integration**: The virtual assistant integrates with ServiceNow to manage support tickets, incidents, and problems, ensuring seamless issue tracking and resolution.
 
 ### Live Agent Integrations
 #### [**Genesys**](/Create/Live%20Agent%20Integrations/Genesys%20Integration)
 - **Genesys Integration**: The virtual assistant integrates with Genesys to provide live agent support, enabling staff to escalate issues to human agents when needed.
 
 ### User Interface Integrations
-- **Microsoft Teams**
+- [**Microsoft Teams**](/Create/User%20Interface%20Integrations/solution-prepare-optional-msteams)
     - Native Integration: The virtual assistant is embedded in Microsoft Teams, providing a seamless and intuitive user experience.
     - Single Platform: Staff can access the virtual assistant directly within Microsoft Teams, reducing the need to switch between platforms.
diff --git a/docs/01-Business Value/02-Use Cases/02-Consumer Assistant.mdx b/docs/01-Business Value/02-Use Cases/02-Consumer Assistant.mdx
@@ -27,9 +27,8 @@ Utilized Watson Discovery to perform RAG document search on the website's data c
 Leveraged wx.ai to provide Spanish translations of user queries and its answers by crawling client's target website (english). 
 
 ### Live Agent Integrations
-#### [**Genesys**](/Create/Live%20Agent%20Integrations/Genesys%20Integration)
-Demonstrated live agent escalation with Genesys from within watsonx Assitant.
+[**Genesys**](/Create/Live%20Agent%20Integrations/Genesys%20Integration):
+The virtual assistant integrates with Genesys to provide live agent support, enabling staff to escalate issues to human agents when needed.
 
 ### User Interface Integrations
-#### [**Hosted Webpage**](/Create/User%20Interface%20Integrations/Hosted%20Static%20Site)
-Adapted the assistant to a custom user-interface
+[**Hosted Static Site**](/Create/User%20Interface%20Integrations/Hosted%20Static%20Site): Adapted the assistant to a custom user-interface
diff --git a/docs/03-Create/03-Multilingual Capability.mdx b/docs/03-Create/03-Multilingual Capability.mdx
@@ -19,7 +19,7 @@ To successfully enable multilingual capabilities we leveraged an LLM via watsonx
 #### **Modify watsonx Assistant actions:**
 1.	Create assistant variable to store selected language by end user.
 2.	Within the "Invoke watsonx generation API" which is configured [here](/Create/solution-create-document%20search#enable-document-search), create a watsonx.ai prompts for step 2 above.
-3. Take translated output from watsonx.ai and pass as the context to carry out the Document search with watson disocvery.
+3. Take translated output from watsonx.ai and pass as the context to carry out the Document search with watson discovery.
 4. Within the "Invoke watsonx generation API" which is configured [here](/Create/solution-create-document%20search#enable-document-search), create a watsonx.ai prompts for step 4 above.
 
 

diff --git a/docs/03-Create/06-Disambiguation Methods.mdx b/docs/03-Create/06-Disambiguation Methods.mdx
@@ -20,6 +20,19 @@ To incorporate a disambiguation method there are two core steps:
     | Printer    | model A, model B      |
 
 2. Build watsonx Assistant actions for each topic and build a step to surface the entities as options to the user
+
     ![](../../assets/screenshots/disam_actions.png)
 
-3. Combine the user's selected option with their document search request and present the clarified answer to them.
+3. Build another step within the watsonx Assistant action to store the selected option as a variable, for example as "selected_option"
+
+    ![](../../assets/screenshots/disam_actions_step.png)
+
+4. Combine the user's "selected option" with the document search request
+
+    For example, if the variable "user_query" holds the user's original question, such as "How do I clear my browser cache?", you'll need to append the selected option to the "user_query" variable before sending the clarified query to the "Document Search" method.
+
+    Example)
+
+    ``` user_query.append(" on").append(selected_option)```
+
+    This will result in the the final query to the "Document Search" method to look like: "How do I clear my browser cache on Safari" which will then return the most appropriate answer.