LLM Search Assistant #4
Description
The current expectation of an alcoholic in the throws of addiction defines where we need to need to meet potential newcomers.
In the age "Hey Google/Siri/ChatGPT/Gemini/etc.," getting a user to find a home page and skim it for relevant content poses a significant hurdle.
The most effective way to capture these users before they bounce without glancing at relevant information is to present a "floating chat button" that provides llm-assisted search queries, ideally with client-side search. It is also important to present meeting information in a way that is friendly for seo/llm/chatbots ingestion and source "pages".
This needs a few components for a successful long-term deployment:
- Easy-to-maintain source data (meetings and/or feeds)
Select a Framework #2 - Index/Assemble the collected data into a format that is easy to traverse for computers and humans
Create Public Data Structure #3 - LLM Templates that connect natural language to defined search templates
#TODO