Ubio
Progress on Previous (Awarded) Submissions
Product & Architecture
Description:
In the ever-evolving landscape of software development, the dependence on AI tools like Github Co-pilot has become indispensable for developers. However, a notable gap exists – these tools are yet to be finely tuned to the unique languages and intricacies of the Stellar blockchain ecosystem. As a passionate AI developer and builder entrenched in the Stellar community, I've identified this challenge and I am excited to propose a solution – the Stellar Developer Co-pilot.
The Problem in Context:
Stellar, with its Soroban-based smart contracts and general Stellar development, has need for a specialized touch that mainstream AI tools lack. The existing solutions fall short when it comes to generating code that is not only syntactically correct but also aligned with the principles and nuances of the Stellar blockchain. This gap poses a bottleneck for developers eager to contribute to the Stellar ecosystem, hindering the speed and efficiency of smart contract development for developers in the ecosystem and creates bigger barriers to entry to new developers
The Solution:
Our proposed Stellar Developer Co-pilot is not just a tool; it's a tailored companion for developers diving into the world of Stellar and Soroban. Picture it as your trusty co-pilot on the coding journey, equipped with a deep understanding of the Stellar ecosystem. This AI assistant is not merely a code generator; it's a learning partner, attuned to the unique requirements of Stellar development.
Why it Matters:
Efficient development is the lifeblood of progress, and with Stellar being at the forefront of blockchain innovation, the need for a dedicated co-pilot is paramount. Closing this gap will empower developers, reduce learning curves, and ultimately accelerate the growth of the Stellar ecosystem. By ensuring that the tools at our disposal align seamlessly with the Stellar languages, we pave the way for a more inclusive and vibrant developer community.
Implementation and Technical Features:
The Stellar Developer Co-pilot will be implemented using state-of-the-art natural language processing (NLP) and machine learning (ML) techniques. Our approach involves training the co-pilot on extensive datasets, ensuring it not only understands the syntax but also captures the essence of Stellar's unique coding practices.
Key technical features include:
- Language Specialization: The co-pilot will be finely tuned to Stellar and Soroban languages, (We will be starting with Soroban first) ensuring that generated code aligns with Stellar's coding standards.
- Context Awareness: The AI will possess a deep understanding of the Stellar ecosystem, allowing it to generate code in context, taking into account the specific requirements of blockchain development.
- Learning and Adaptation: The co-pilot will continually learn from user interactions, evolving to meet the evolving needs of the Stellar developer community.
Tech Stack:
The technology stack will leverage cutting-edge NLP libraries, machine learning frameworks, and the robust infrastructure necessary for the development of a sophisticated AI tool. I am committed to transparency and collaboration, welcoming community input to refine and enhance the Stellar Developer Co-pilot.
In essence, this proposal seeks not just to address a technical gap but to foster a more collaborative, efficient, and empowered Stellar developer community. Together, let's embark on a coding journey where the co-pilot is as passionate about Stellar as the developers themselves.
USAGE:
The solution will be open-sourced so other developers in the ecosystem can adapt it in any format suitable to them, but for us we will make a simple interface where developers can simply access the solution, we will make the API available to the community and we will also develop it as a plugin on Chatgpt so as to reach a larger audience
IMPACT:
The proposed Stellar Developer Co-pilot isn't just another fancy tool; it's a game-changer for the community. Picture this: developers diving into Soroban and Stellar with newfound ease, confidently crafting smart contracts like it's second nature. Now that's the kind of value we're talking about.
Empowering Developers:
Imagine a world where developers, seasoned or just starting out, feel like they have a buddy in the coding trenches. Our co-pilot isn't just about generating code; it's about empowering developers to contribute effectively to the Stellar ecosystem. With a specialized touch for Stellar languages, the learning curve gets a bit less steep, and the possibilities for innovation broaden.
Accelerating Development:
Faster, more efficient coding means projects move at warp speed. The success of our co-pilot translates to quicker smart contract development. That's not just a win for individual devs; it's a boost for the entire Stellar community. The more we can streamline the development process, the faster we can bring exciting projects to life.
Fostering Collaboration:
You know what happens when tools are tailored to the community? Collaboration skyrockets. Our co-pilot isn't an outsider; it's part of the Stellar crew. By aligning with Stellar's coding practices, it becomes a seamless addition to the developer toolkit, encouraging collaboration and knowledge-sharing within the community.
Wider Reach
Our plan is to make this solution easily accessible to all and to make its API available to all and the plugins integrated into ChatGPT plugins, this will bring new reach to the Stellar developer community as more people begin to play with the tool to learn how to build their first Stellar Dapps and Soroban Smart Contracts
Deliverables List
Description: In this stage, we will be gathering all the relevant data, info, and codes from all the repositories and sources available, and making it suitable for data processing and training of our StellarGPT.
How to Measure Completion:
(I)A public repository is created on GitHub for compiling all the cleansed data
(ii) The repository contains cleansed and curated data of hundreds of Stellar and Soroban codes
(iii) The repository contains hundreds of curated data of Stellar and Soroban info
Estimated Date: 2 Weeks
Budget: $10,000
Deliverable 2: Initial Model Build and Finetuning
Description: Processing, training, fine-tuning, and testing of the model using a large language model
How to Measure Completion:
(I) A code base for the functional fine-tuned model
(ii) The model can generate relative Stellar and Soroban codes and info
(iii) A video demo of the model performing the functionality described in (ii)
Estimated Date: 6 Weeks
Budget: $35000
Pitch Deck URL
Project Categories
Product Type [If Other Developer Tooling]
Section
Section
Technical Architecture Doc
Project URL
Code URL
Video URL
Requested Budget in USD ($)
Total Expected Roadmap
(i) Improvement of model accuracy, results, contexts, and outputs
Budget: $20,000
(ii) API and SDK build for the model
Budget: $10,000
(iii) Front end Build and Integration into Front-end
Budget: $15,000
(iv) Integration as a ChatGPT plugin
Budget: $15,000
(v) Integration as a Chrome Extension
Budget: $15,000
Section
Team bio
UBIO OBU: The CEO of Remostart, is a blockchain and AI developer and researcher, his academic background is in Artificial Intelligence. Ubio has about 4 years of experience in researching, his research works have cut across different fields, from AI, to IoT, Agriculture, environment, blockchain, HR, human behaviors etc. He currently has about 7 research paper publications in reputable journals like the American Institute of Physics, and IEEE, etc, he has 2 patents under application and a Copyright on a book titled "Research writing for beginners".
Ubio is a Microsoft Winsider Recipient, KECTIL leadership fellow, and a SingularityNet Ambassador.
He has built several AI models and led the team that built the Virtual farm, an AI based Computer vision farm.
Ubio will be in charge of the AI development and resource management
https://www.linkedin.com/in/ubio-obu-71927276/
Ediyangha Otogho: Full-stack software and Blockchain developer with 8 years of of software development experience and 3 years blockchain development experience. Ediyangha has won several hackathons and techatrons and was the chief technology officer behind Send funds, a fintech solution building a Bharatpe for Africa. Funfact Ediyangha can code efficiently in more than 7 programming languages. For this project he will be the blockchain and fullstack developer
https://www.linkedin.com/in/edinyanga-ottoho-02801517a/
https://www.github.com/EdinyangaOttoho
Yash Ambekar : B.Tech-Computer Engineering, Full stack Developer, 7 years experience in Software development, a Smart India hackathon winner, with about 3 Research paper publications. In this project he will coordinate the front end developer especially the UI/UX aspects.
Daniel Effiom: He is a co-founder at RemoStart, a Reconciliation analyst at ETransact international PLC. With 5 years experience in data analysis, process monitoring and operational procedures. He has managed several projects for RemoStart and ETransact and is why he will be the project and product manager for this project.