Python/LLaMA Tech-Support Chatbot
Budget: €750 – €1,500 EUR
Python/LLaMA Tech-Support Chatbot
This document outlines the specifications for a Python-based tech-support chatbot using the LLaMA language model. The chatbot is intended to provide efficient responses to customer queries, leveraging user-uploaded manuals for contextually accurate answers.
= Requirements =
* Written in Python 3.
* Runs as a WSGI application under an Apache server.
* Utilizes the LLaMA language model for natural language understanding and responses.
-------------------
= Public chat UI =
Allows users to:
* Ask multiple questions in a conversational chat format.
* Refine or clarify their questions iteratively.
-------------------
= Admin UI =
* Provides an interface to upload and manage plain-text and HTML-based user manuals.
* Enables the chatbot to integrate these manuals for context-aware responses.
-------------------
= Constraints =
* The chatbot must operate entirely within the server environment.
* No external API calls or dependencies on third-party services.
-------------------
= Functional specifications =
Backend:
* Configure and fine-tune the LLaMA model to answer tech-support queries and cross-reference uploaded user manuals.
* Extract searchable text and metadata from the uploaded plain-text and HTML manuals.
* Index the content for efficient querying by the chatbot.
* Use the LLaMA model to understand user intent and provide accurate responses, combining model knowledge with manual content.
* Support multi-turn conversations and maintain state to understand context across questions in a session.
User chat UI:
* Design a simple, intuitive web interface.
* Display a conversation history for user reference.
* Enable input refinement with minimal friction.
Admin UI:
* Include file upload functionality for manuals.
* Provide feedback on the parsing status (e.g., success or errors).
* Offer a dashboard to view and manage uploaded content.
Security:
* Ensure user data and chat sessions are sandboxed.
* Validate all inputs to prevent injections or other attacks.
* Restrict admin interface access to authorized users.
* Enforce secure authentication mechanisms.
* Implement robust error detection and graceful recovery mechanisms for both frontend and backend.
Scalability:
* Architect the system for handling multiple concurrent users without performance degradation.
* Design with modular components to allow future enhancements (e.g., adding new data sources or support for additional languages).
Deliverables:
* Well-documented Python scripts for the chatbot and its integration with the LLaMA model.
* Step-by-step instructions for setting up the application on an Apache server with WSGI.
* Include test scripts for validating core functionalities and ensuring robustness.
This document outlines the specifications for a Python-based tech-support chatbot using the LLaMA language model. The chatbot is intended to provide efficient responses to customer queries, leveraging user-uploaded manuals for contextually accurate answers.
= Requirements =
* Written in Python 3.
* Runs as a WSGI application under an Apache server.
* Utilizes the LLaMA language model for natural language understanding and responses.
-------------------
= Public chat UI =
Allows users to:
* Ask multiple questions in a conversational chat format.
* Refine or clarify their questions iteratively.
-------------------
= Admin UI =
* Provides an interface to upload and manage plain-text and HTML-based user manuals.
* Enables the chatbot to integrate these manuals for context-aware responses.
-------------------
= Constraints =
* The chatbot must operate entirely within the server environment.
* No external API calls or dependencies on third-party services.
-------------------
= Functional specifications =
Backend:
* Configure and fine-tune the LLaMA model to answer tech-support queries and cross-reference uploaded user manuals.
* Extract searchable text and metadata from the uploaded plain-text and HTML manuals.
* Index the content for efficient querying by the chatbot.
* Use the LLaMA model to understand user intent and provide accurate responses, combining model knowledge with manual content.
* Support multi-turn conversations and maintain state to understand context across questions in a session.
User chat UI:
* Design a simple, intuitive web interface.
* Display a conversation history for user reference.
* Enable input refinement with minimal friction.
Admin UI:
* Include file upload functionality for manuals.
* Provide feedback on the parsing status (e.g., success or errors).
* Offer a dashboard to view and manage uploaded content.
Security:
* Ensure user data and chat sessions are sandboxed.
* Validate all inputs to prevent injections or other attacks.
* Restrict admin interface access to authorized users.
* Enforce secure authentication mechanisms.
* Implement robust error detection and graceful recovery mechanisms for both frontend and backend.
Scalability:
* Architect the system for handling multiple concurrent users without performance degradation.
* Design with modular components to allow future enhancements (e.g., adding new data sources or support for additional languages).
Deliverables:
* Well-documented Python scripts for the chatbot and its integration with the LLaMA model.
* Step-by-step instructions for setting up the application on an Apache server with WSGI.
* Include test scripts for validating core functionalities and ensuring robustness.