Implementation of letmeexplain.ai

[letmeexplainAI]

This PR integrates the LetMeExplain.ai chatbot into the Holochain developer docs to provide AI-powered responses to user queries. The chatbot is loaded via an external script and connects to a Large Language Model (LLM) to generate responses. The LLM is currently trained only on the Holochain Whitepaper.

Technical Details:

• The script (lme_chatbot_widget.js) is hosted on https://dashboard.letmeexplain.ai.
• It initializes a chatbot widget with a unique orgId to identify the client.
• The chatbot interacts with an LLM to respond to user queries.

Security and Privacy:

• The script does not collect or store sensitive user data.

Support:

For access to the LetMeExplain.ai client dashboard or technical support, please contact:

• Email: hello@letmeexplain.ai
• Discord: @minus1618
• Telegram: @amiri_uk

[pdaoust]

Hey @letmeexplainAI , interesting proposal, and it's something we've contemplated adding to our dev docs already.

Right now our first priority is getting actual documentation written, then training an LLM on it. The white paper already reflects a slightly out-of-date model of the living code, and will get more stale as time goes on. Can you tell me more about your product? Big questions are:

1. Will it progressively update its model as our documentation improves? (I assume it's doing RAG or fine-tuning or building up a huge context window or something)
2. What base model does it use?
3. What are your pricing plans for open-source projects stewarded by non-profits like ours?

[letmeexplainAI]

Hi, Thank you for reaching out. Below are the answers to your questions regarding our product. 1. *Progressive Model Updates:* Yes, LME is built on a *RAG (Retrieval-Augmented Generation)* architecture using open-source technologies on AWS. Documentation updates can be managed through the client dashboard, where you can manually add new content. Additionally, we periodically scan existing URLs to ensure your knowledge base remains up to date. 2. *Base Model:* We use *OpenAI's 4o-mini* as the foundational model for LME. 3. *Pricing for Non-Profits: *We are an AI Lab and love nonprofits. - The *whitepaper-based chatbot and general Q&A features* are available for free. - If your documentation dataset expands significantly or user traffic increases, you have the option to *continue using the free bot on the whitepaper* or transition to an affordable paid plan that primarily covers our operational costs. Please let us know if you have any further questions or if you'd like a demo. Thanks, Owais

…

On Wed, Feb 5, 2025 at 7:18 PM pdaoust ***@***.***> wrote: Hey @letmeexplainAI <https://github.com/letmeexplainAI> , interesting proposal, and it's something we've contemplated adding to our dev docs already. Right now our first priority is getting actual documentation written, then training an LLM on it. The white paper already reflects a slightly out-of-date model of the living code, and will get more stale as time goes on. Can you tell me more about your product? Big questions are: 1. Will it progressively update its model as our documentation improves? (I assume it's doing RAG or fine-tuning or building up a huge context window or something) 2. What base model does it use? 3. What are your pricing plans for open-source projects stewarded by non-profits like ours? — Reply to this email directly, view it on GitHub <#521 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BPAVMTZCGYKWTSHMZUNKJ332OJPZNAVCNFSM6AAAAABWN3L4T2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZXHAZDINBRHE> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[letmeexplainAI]

Hi there, Following up on the previous email, I hope the responses provided addressed your questions adequately. Please let me know if you have any further questions or require additional information. I'd be happy to schedule a call/demo at your convenience. Best Regards, Owais On Fri, Feb 7, 2025 at 9:37 AM LetMeExplain AI ***@***.***> wrote:

…

Hi, Thank you for reaching out. Below are the answers to your questions regarding our product. 1. *Progressive Model Updates:* Yes, LME is built on a *RAG (Retrieval-Augmented Generation)* architecture using open-source technologies on AWS. Documentation updates can be managed through the client dashboard, where you can manually add new content. Additionally, we periodically scan existing URLs to ensure your knowledge base remains up to date. 2. *Base Model:* We use *OpenAI's 4o-mini* as the foundational model for LME. 3. *Pricing for Non-Profits: *We are an AI Lab and love nonprofits. - The *whitepaper-based chatbot and general Q&A features* are available for free. - If your documentation dataset expands significantly or user traffic increases, you have the option to *continue using the free bot on the whitepaper* or transition to an affordable paid plan that primarily covers our operational costs. Please let us know if you have any further questions or if you'd like a demo. Thanks, Owais On Wed, Feb 5, 2025 at 7:18 PM pdaoust ***@***.***> wrote: > Hey @letmeexplainAI <https://github.com/letmeexplainAI> , interesting > proposal, and it's something we've contemplated adding to our dev docs > already. > > Right now our first priority is getting actual documentation written, > then training an LLM on it. The white paper already reflects a slightly > out-of-date model of the living code, and will get more stale as time goes > on. Can you tell me more about your product? Big questions are: > > 1. Will it progressively update its model as our documentation > improves? (I assume it's doing RAG or fine-tuning or building up a huge > context window or something) > 2. What base model does it use? > 3. What are your pricing plans for open-source projects stewarded by > non-profits like ours? > > — > Reply to this email directly, view it on GitHub > <#521 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/BPAVMTZCGYKWTSHMZUNKJ332OJPZNAVCNFSM6AAAAABWN3L4T2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZXHAZDINBRHE> > . > You are receiving this because you were mentioned.Message ID: > ***@***.***> >

[pdaoust]

Thanks! Every question answered except one: what are your pricing plans? I don't see that on your website.

I'd also like to know: how long have you been in business?

Thanks!