ChatGPT Plugin fine tuned with Atala PRISM Documentation, OpenAPI Spec and W3C standards.

[GENERAL] Name and surname of main applicant

Roberto Carvajal

[GENERAL] Email address of main applicant

roberto.carvajal@gmail.com

Additional applicants

Jon Bauer

[GENERAL] Please specify how many months you expect your project to last (from 2-12 months)

4

[GENERAL] Please indicate if your proposal has been auto-translated into English from another language.

No

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

Yes

[GENERAL] If YES, please describe what the dependency is and why you believe it is essential for your project’s delivery. If NO, please write “No dependencies.” .

The project will run inside the ChatGPT interface and loaded as a plugin.

[GENERAL] Will your project’s output/s be fully open source?

Yes

[GENERAL] If NO, please describe which outputs are not going to be open source. If YES, please write “Project will be fully open source.”

Project will be fully open source.

[METADATA] Category of proposal

Education

[IMPACT] Please describe your proposed solution.

Generative AI can be to be used as a tool for education in the form of a tutor to learn about any topic. The ability to "have a conversation" around a topic is a natural way for people to learn and explore new information.

In particular, SSI concepts and decentralization are hard to grasp for anyone at the start of their journey and having a companion "chat bot" available 24x7 with deep knowledge in these topics that can help you not only solve questions and solidify novel concepts but also produce working code in your preferred language is a formidable tool for anyone in the community.

[IMPACT] How does your proposed solution address the challenge and what benefits will this bring to the Cardano ecosystem?

This project has the capacity to help reduce the entry barriers in understanding SSI concepts and also help anyone kickstart their PRISM projects by providing working code examples in the preferred language of the user, this is helpful for both novice and expert programmers and even non technical people exploring SSI and PRISM. This powerful tool trained to understand JSON formatted structures for Verifiable Credentials can help anyone designing these interactions and quickly test ideas and enable rapid prototyping - thus accelerating the adoption of Atala PRISM.

[IMPACT] How do you intend to measure the success of your project?

This innovation will directly improve the productivity of anyone building and prototyping solutions using Atala PRISM. Success will be measured in the usage metrics of the plugin itself.

[IMPACT] Please describe your plans to share the outputs and results of your project?

The project is fully open source and the process of building the plugin will be documented in every step of the way, anyone is welcome to contribute and also anyone could use it as a template to learn how to build ChatGPT plugins in general as the steps are reproducible to be applied to any topic.

Any script or tool used to achieve the creation of the plugin will be documented or will be already an open source project, e.g. the index of the data we want to use for training, the crawlers to fetch the documentation, libraries to split the content in chunks for training, formatting data, generating library API clients for Atala PRISM with OpenAPI, links to resources, etc, the idea is to fully document the process of how the plugin is created, be completely transparent on what data is used for training and how to setup the actual plugin manifest and publish it on the OpenAI plugin repository to be used by the community.

[CAPABILITY/ FEASIBILITY] What is your capability to deliver your project with high levels of trust and accountability?

I am a Senior Full Stack Software Developer with 20 years of experience with web and mobile technologies and have fulfilled various roles across the years from sysadmin, backend & frontend dev and architect. I am currently exploring SSI based solutions and have recently taken all available Pioneer program courses (AP101-AP102-AP103). I have worked over the years in many projects involved crawling, data management and generally building APIs. The technical challenge of connecting to ChatGPT API and fine-tuning a model fits comfortably within my skillsets.

[CAPABILITY/ FEASIBILITY] What are the main goals for the project and how will you validate if your approach is feasible?

This project have a high level of feasibility, the real challenge is mostly deciding what data and what format is better suited to achieve the best results. For this reason the open source nature of the project is critical so it can receive feedback from the community, proposing the right technical papers to train on, example Verifiable Credentials, etc.

[CAPABILITY/ FEASIBILITY] Please provide a detailed breakdown of your project’s milestones and each of the main tasks or activities to reach the milestone plus the expected timeline for the delivery.

• M1 Initial research and callout to the community to propose the best technical documentation available to train and fine tune the model.
• M2 Creation of tooling (scripts, crawlers, formatters, etc)
• M3 Indexing and preparing data for training
• M4 Documentation (written as the project moves forward)
• M5 Training the model
• M6 Publishing the plugin in OpenAI repository

[CAPABILITY/ FEASIBILITY] Please describe the deliverables, outputs and intended outcomes of each milestone.

• M1 GitHub repository with the starting list of links of data to be indexed
• M2 Scripts and documentation on how to fetch the data and format it for training
• M3 The actual formatted data ready for training
• M4 Documentation (this is likely created on the go, but I want to have a clear milestone)
• M5 The model trained and available in OpenAI
• M6 The plugin available for anyone using ChatGPT (inside OpenAI plugin repository)

[RESOURCES & VALUE FOR MONEY] Please provide a detailed budget breakdown of the proposed work and resources.

• OpenAI API Pricing cost will depend on how much data is trained, ballpark of maybe ₳8,000
• Linode server instance, ₳2000 (4 months)
• Development labor, research, etc ₳50,000 (4 months)

[RESOURCES & VALUE FOR MONEY] How does the cost of the project represent value for money for the Cardano ecosystem?

Cost is estimated on time dedicated as senior developer for the creation of tooling and documentation to support formatted data to be usable for training in larga language models such as ChatGPT.

[IMPORTANT NOTE] The Applicant agreed to Fund10 rules and also that data in the Submission Form and other data provided by the project team during the course of the project will be publicly available.

I Accept