Precisely, I use AI everyday for hours on end lol, its helped me tremendously formulating protocols for myself. And yes Fullscript is the bomb, got a whole panel on myself the other week costing me 120 bucks. If I had been billed via insurance, I probably would have payed 400!
Great question! I am sure you are familiar with RAG? Essentially, I have a database filled with contextual information, person asks a question, hits a similarity score in the database, returns the information for the AI to parse through. Now, like you said, individuals are going to have multiple variables according to their own health status. Its not going to be perfect, but having Maximal Marginal Relevance and hybrid search built in will help with this. Say a query involves mold toxicity, or auotimmune, or something to that affect, its going to try and pull a variation of data relevant to that. Someone with mold might have impaired methylation (almost always actually haha), their gut microbiome might be off, their cognition might be hindered, etc. Will it get it every time, no. We have the ability to think horizontally, as I like to think, AI is best at addressing whatever is right in front of it. Not sure if that makes sense? Almost like thinking in layers. However, with all AI the power really comes from someone who can prompt effectively, so educating people on how to use it to maximize it will be a challenge! Not sure if that helped at all, let me know! Also I would be curious if you had any fun ideas in how to implement agents into a health coaching business!
Hey thanks a lot, appreciate that!
Precisely, I use AI everyday for hours on end lol, its helped me tremendously formulating protocols for myself. And yes Fullscript is the bomb, got a whole panel on myself the other week costing me 120 bucks. If I had been billed via insurance, I probably would have payed 400!
Great question! I am sure you are familiar with RAG? Essentially, I have a database filled with contextual information, person asks a question, hits a similarity score in the database, returns the information for the AI to parse through. Now, like you said, individuals are going to have multiple variables according to their own health status. Its not going to be perfect, but having Maximal Marginal Relevance and hybrid search built in will help with this. Say a query involves mold toxicity, or auotimmune, or something to that affect, its going to try and pull a variation of data relevant to that. Someone with mold might have impaired methylation (almost always actually haha), their gut microbiome might be off, their cognition might be hindered, etc. Will it get it every time, no. We have the ability to think horizontally, as I like to think, AI is best at addressing whatever is right in front of it. Not sure if that makes sense? Almost like thinking in layers. However, with all AI the power really comes from someone who can prompt effectively, so educating people on how to use it to maximize it will be a challenge! Not sure if that helped at all, let me know! Also I would be curious if you had any fun ideas in how to implement agents into a health coaching business!