AI Colleague Your Project’s Institutional Memory

[Proposal setup] Proposal title

Please provide your proposal title

AI Colleague Your Project’s Institutional Memory

[Proposal Summary] Budget Information

Enter the amount of funding you are requesting in ADA

90000

[Proposal Summary] Time

Please specify how many months you expect your project to last

6

[Proposal Summary] Translation Information

Please indicate if your proposal has been auto-translated

No

Original Language

en

[Proposal Summary] Problem Statement

What is the problem you want to solve?

Complex development cycle for Cardano dApps and projects which prvents more to be launched on the entwork. Such as different languages for on-chain and off-chain and development enviroment set up.

[Proposal Summary] Supporting Documentation

Supporting links

• https://github.com/ricardas-softa/cardano-smart

[Proposal Summary] Project Dependencies

Does your project have any dependencies on other organizations, technical or otherwise?

Yes

Describe any dependencies or write 'No dependencies'

LLM models that can be self hosted: Llama, Mystral, DeepSeek, etc.

[Proposal Summary] Project Open Source

Will your project's outputs be fully open source?

Yes

License and Additional Information

It'll be released using an the MIT License. The idea is to release a virtual assistant that any project can clone and use for their own particular needs. The various version can be kept private or released for others to use in a collaborative enviroment.

[Theme Selection] Theme

Please choose the most relevant theme and tag related to the outcomes of your proposal.

AI

[Campaign Category] Category Questions

Describe what makes your idea innovative compared to what has been previously funded (whether by you or others).

Previous iterations of Cardano AI tools relied on paid subscription models such as ChatGPT or Claude. These are API based and constantly need more payments to keep themselves operational. Our solution uses LLM models that can be self-hosted by any person who wants to use them.

The self-hosted models do not share data with their parent developers, meaning they can keep information private for those startups developing on Cardano who want to remain private before launch.

Describe what your prototype or MVP will demonstrate, and where it can be accessed.

The product will be a complete AI assistant that can be launched on a home computer, private server, or cloud environment. It won't require any subscriptions, nor will there be any paywalls that block use of the tool by anyone. Later, we plan to launch this as a complete product that we'll host a version with a subscription model. But this launch is proof that LLM models can speed Cardano development and assist teams with their projects.

Describe realistic measures of success, ideally with on-chain metrics.

The measure we'll use primarily will be how many examples of code the LLM model can produce and how many Cardano questions it can answer. Additionally, there will be the number of times the code repository has been cloned by the various people who access it. The repository will contain options of different models that can be used from home hardware to cloud computing.

[Your Project and Solution] Solution

Please describe your proposed solution and how it addresses the problem

“AI Colleague” analyzes the entire codebase and related files, indexing them into an optimized knowledge store. It acts like an intelligent chatbot and documentation generator:

• Answer any question about the codebase or project.
• Generate Markdown or PDF documentation with structure and context.
• Efficiently caches analyzed files to avoid reprocessing unchanged ones.
• Feels like chatting with a senior dev who built the system years ago.

The Bot is also deployable as a chat interface on the main site of a project. It'll answer any questions users may have, like how to stake a token, how to create a liquidity pool, how to borrow an asset, etc. This means any Cardano project will have access to a tool that helps them with development, documentation, and user outreach.

[Your Project and Solution] Impact

Please define the positive impact your project will have on the wider Cardano community

Cardano remains a chain that is not considered "friendly" by most web3 developers. This is mainly because it differs greatly from the account model of other blockchains. New development languages and projects have addressed this hurdle, but outside Cardano, these are unknown tools.

The AI assistant can help any new developer to quickly find the best tools to create Cardano-based dApps. Also, it can help users navigate the offereings of a new product and see how it works for maximum benefit.

[Your Project and Solution] Capabilities & 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?

As a team, we have worked in Cardano for several years. We launched a project on-chain that included tokenization and supply chain traceability that used the Cardano ledger. We have created AI tools for image generation, code review, and chatbots for several types of industries. We worked with both API private models and also open-source models that can be self-hosted.

[Milestones] Project Milestones

Milestone Title

Setup of the project

Milestone Outputs

• Initial setup of the modules for the project, including the LLM models.
• Creation of an interface to interact with the bot.
• Creation of a code repository for the project that is open and public, so anyone can track its progress.

Acceptance Criteria

There is a code repository where all the modules are located and populated by their respective roles. There is an interface that is included in the repository that, when cloned, can be used to interact with the bot. There is a set of LLM models that can be chosen with the interface to maximize the hardware resources of the user.

Evidence of Completion

• GitHub repository with all the modules open.
• There is an interface added in the repo that can be used to interact with the project.
• There are at least three LLM model choices that can be selected depending on the hardware resources.

Delivery Month

2

Cost

30000

Progress

30 %

Milestone Title

Deployment and testing

Milestone Outputs

A fully deployable bot that, when cloned from the repo, deploys in the local environment.

The bot only requires a few configuration choices, such as the type of LLM to be used after the bot deploys without the need to modify the code.

There are benchmarks that show how the different model choices work with hardware environments, from local setups using commercial hardware to cloud and server choices.

Acceptance Criteria

There is a repository that deploys both by just making a few configuration choices that set up the LLM. There is a part that suggests which model is optimal depending on the hardware used for the bot. There are examples of the bot performing with different model configurations that are provided in video form.

Evidence of Completion

GitHub repository with the bot and the configuration choices.

Benchmarks for the hardware types.

Videos with the different model choices performing various tasks.

Delivery Month

4

Cost

30000

Progress

70 %

Milestone Title

Optimizing and Outreach

Milestone Outputs

Optimize the bots to run in various environments.

Create an API that can deploy the interface on online pages to serve as a chatbot.

Showcase the project in various places to reach the intended user base.

Acceptance Criteria

The code repository has all the necessary documentation to deploy the bot in various environments and for different uses. There is an API for web pages to act as a chatbot. There are instructional videos showing the deployment of the tools in its various forms and for different use cases.

Evidence of Completion

Complete GitHub repository with LLM options and documentation.

API for the bot to connect to a webpage.

Instructional videos on how best to deploy the bot, or what types of hardware, and the different use cases for the tool.

Delivery Month

6

Cost

30000

Progress

100 %

[Final Pitch] Budget & Costs

Please provide a cost breakdown of the proposed work and resources

Setup 30000 ADA

Creation of GitHub repo	

Initial setup of the tool	

Selection of LLMs	

Creation of the interface	

Deployment 30000 ADA

Deploy the bot on various hardware		

Showcase of the different LLM choices		

Optimization of the tool		

Active Support 30000 ADA

GitHub support and Documentation		

Community engagement on X and Discord		

Showcase of the API

[Final Pitch] Value for Money

How does the cost of the project represent value for the Cardano ecosystem?

The community will benefit from a tool that can help developers and Cardano users. It'll work as an assistant during the creation of a new dApp or product, while once it is released, it'll act as a chatbot that can guide users through the platform. The tool will be self-hosted, so there won't be any need for paid subscriptions, and also free to deploy on various websites as a chatbot.

[Required Acknowledgements] Consent & Confirmation

Terms and Conditions:

Yes