Closed NicolasZoellner closed 1 year ago
Develop a concept about:
Criterias:
TLDR: Here is a "Playground" of an already fully working sample using the current docs based on the Casper Docs GitHub Repo: https://markprompt.com/s/EGKw9g8t
I think with this very small budget this is a lot what you get already!
I talked to another developer who did build an AI integration for another blockchain project which is also based on Docusaurus (IIRC) but this is all programmed manually and specific to the project. That is done in his free time - otherwise this would be far too expensive. For now his solution is also not portable to other projects.
The main difference is that with a custom solution you can e. g. ask the current block height or other custom and dynamic things but that is not needed nor useful for now and needs custom programming and ongoing maintenance.
I'm sure there is more to add here but I would prefer to talk to someone who can answer some questions or even has questions which I can answer 🙂
The concept was introduced by GuybrushX on 22.06.23 to the Casper Association, and all open questions could have been answered. The concept is now completed and will be taken into further consideration within the Casper Association. Therefore, the DevReward will be changed from Open to Completed. Thank you for your participation, GuybrushX!
Dev Reward Title
Prepare a general Concept for an Integration of an AI support system with Casper Developer Portal
Reward Size
350 USD
Reward category
Documentation
Description
The main objective of this DevReward is to develop a comprehensive concept for integrating GPT-4 or a comparable AI model as a support tool for the Casper Developer Portal (Casper.network) by utilizing all available internal data.
The expected concept includes the following steps, which should be considered in terms of budgeting and planning:
Gather internal data: Collect all relevant internal data from various sources such as FAQs, knowledge base articles, and historical customer interactions available on the Casper Developer Portal.
Preprocess and format the data: Clean and format the internal data to ensure its suitability for training GPT-4. This may involve removing irrelevant information, standardizing the format, and organizing it in a structured manner.
Train AI Model: Utilize the preprocessed internal data to train an AI model using an appropriate training pipeline. This step involves feeding the data into the model and allowing it to learn the patterns, context, and nuances specific to the Casper Developer Portal.
Develop a user interface: Create a user interface that enables seamless interaction with the AI model. This can take the form of a chatbot or a search function embedded within the website.
Implement real-time data updates: Set up a system to regularly update the AI model with the latest internal data. This ensures that the model remains up to date and can provide accurate responses based on the most recent information available.
Test and refine: Conduct thorough testing by simulating user interactions and evaluating the responses provided by the AI model. Identify areas for improvement or refinement and iterate on the system accordingly.
Deploy and monitor: Once the integration has been thoroughly tested and refined, deploy the AI model support tool on the Casper Developer Portal. Continuously monitor its performance, gather user feedback, and make necessary adjustments to enhance its effectiveness and usability.
The acceptance criteria for the AI integration in the Casper Developer Portal are as follows:
Accuracy of Responses: The chatbot should provide accurate and relevant responses to user queries, which will be evaluated by internal Developer Advocates.
Understanding User Intent: The chatbot should accurately interpret user intent and respond accordingly. It should comprehend the nuances and context of user queries to provide appropriate and helpful responses.
Response Time: The chatbot should respond to user queries within an acceptable timeframe, as determined by Developer Advocates.
Language Comprehension: The chatbot should have a wide vocabulary and understanding of various language nuances to facilitate effective communication.
Handling of User Errors: The chatbot should be capable of effectively handling user errors and misunderstandings. It should provide clear prompts or suggestions when a user query is ambiguous or invalid, guiding them towards the correct information.
Personalization and Context Persistence: The chatbot should be able to maintain context throughout the conversation and provide personalized responses based on user preferences or previous interactions.
Integration with Existing Systems: The integration should successfully connect with all relevant data sources, such as the Developer portal and CasperEcosystem.io, ensuring seamless data exchange.
Scalability and Performance: The chatbot should be able to handle a high volume of concurrent user interactions without a significant decrease in performance. Performance benchmarks may be defined as part of the acceptance criteria to ensure effective scalability.
Continuous Improvement: The acceptance criteria should include provisions for ongoing monitoring and improvement of the chatbot's performance based on user feedback and analytics data.
Acceptance Criteria
To develop a comprehensive concept for the desired outcome, the following tasks should be included: