[SOLUTION] Please describe your proposed solution.
The Cardano ecosystem needs a concise and fast way to answer development questions. The best solution with the current technology is to leverage the power of Large Language Models (LLM) and vectorial databases.
Cardano Smart is a platform where a script scrapes the official repositories of all the major Cardano developer tools, incorporates them into a vectorial database, tests a series of LLM models, and finally chooses the optimal one.
Once the process finishes, the users can ask prompts to the model. It, in turn, queries the data in the system and gives the best answers it’s able to find. All the answers are sourced so users can verify the authenticity of the information.
The current Cardano Smart bot can answer questions from the Cardano papers. Additionally, it can provide code for PlutusTx, Aiken, Lucid, and Haskell. We aim to add more, we’re currently looking to add Mesh, Helios, and Plu-ts.
We’re currently doing an NFT sale that will also contribute to the project:
[IMPACT] Please define the positive impact your project will have on the wider Cardano community.
As we stated, the Cardano documentation is fragmented across many repositories, sites, and examples. For new developers, it’s hard to get this info as they need to search the web, ask in Discord servers, read outdated documents, etc.
The Cardano Smart bot is able to amalgamate all this scattered information in one tool. The bot also works great for learning as it provides complete answers and provides sources. The person asking can learn new things just by asking questions on the same topic or reading through the sources.
[CAPABILITY & FEASIBILITY] What is your capability to deliver your project with high levels of trust and accountability? How do you intend to validate if your approach is feasible?
We already have a working demo of the prototype as people can see on the provided YouTube video. All the developers needed to make this happen are currently part of the project, so there is no need to subcontract any more people.
The current demo runs on a local computer. As we need funds to rent server space with GPUs to speed up the bot and train more powerful models.
[Project Milestones] What are the key milestones you need to achieve in order to complete your project successfully?
Run the Cardano Smart prototype from a Server.
Code an API for easy access.
Beta test the bot using people accessing it at the same time from remote locations.
The final product would be a complete version of Cardano Smart running on servers that is able to answer questions about Cardano theory. It will also give code examples for PlutusTx, Aiken, Lucid, and Mesh.
It’s possible more Cardano languages could be added during the development. It depends on our resources.
[RESOURCES] Who is in the project team and what are their roles?
Ričardas Darkšas lead engineer. In charge of loading the material to the vectorial database and testing different models for the Cardano Smart bot.
More information to come.
[BUDGET & COSTS] Please provide a cost breakdown of the proposed work and resources.
Introduce a breakdown of server costs and developers’ paid time.
[VALUE FOR MONEY] How does the cost of the project represent value for money for the Cardano ecosystem?
The Cardano Smart bot will increase the pace of development for Cardano. The bot will provide on-demand and up-to-date information for developers, students, and the general public. It will speed up development times and also help those just entering the ecosystem.
We aim to make it a one-stop shop for all the information regarding Cardano and the entire ecosystem.