Boris Boarman: AI-assistant helping navigate crypto grants

4mos ago
0 Comments

This post is for initial discussion and feedback collection on an AI-assistant designed to help navigate crypto grants, bounties, and other funding options. The full treasury proposal is in progress and will be published and shared as soon as possible. Below, you can find the Executive Summary, Problem Statement, and Project Objectives to familiarize the community with the idea and improve the proposal for success.

Context & Objective

The Web3 space is well-known for fostering innovation, and existing crypto grant programs and funding options have given birth to many promising startups and initiatives. However, the process of searching for, finding, and applying for the right crypto funding option is cumbersome and time-consuming for both applicants and reviewers. Applicants struggle with identifying suitable grants and preparing high-quality applications, while reviewers face the challenge of managing numerous submissions. This inefficiency leads to missed funding opportunities for promising projects and increased workloads for grant reviewers.

Solution: Boris Boarman

Having experienced this struggle ourselves, we conceived Boris Boarman, a conversational AI designed to streamline the application process for funds seekers and crypto foundations. By leveraging advanced AI technology, the platform assists users in validating their project ideas, identifying suitable grant opportunities, and generating high-quality application documents, including spreadsheets with necessary calculations and pre-filled application information.

Our research identified the same problem among many other crypto foundations and similar projects. However, compared to the alternatives, Boris Boarman offers not just fast searching assistance but real artificial intelligence for validating ideas and automating manual labor. This allows founders and active community users to focus on project delivery instead of bureaucratic work.

Use Cases
Our research identified three scenarios where such an AI assistant can be helpful:

Epic 1: Idea Review and Evaluation

  • An idea is checked against the knowledge base of all previous applications and tested by a trained LLM or real AI capable of evaluating an idea based on a numeric impact analysis approach. Feedback for improvement is provided.

Epic 2: Finding Relevant Grant Programs

  • The AI assistant searches for and finds the best and currently active funding options to finance the development of the project or initiative, reducing the effort from days to minutes.

Epic 3: Application Document Generation

  • The AI assistant guides the user through a sequence of questions to collect enough information about the project and generates application documents (including the application form and project development calculation) following a foundation's application structure.

Advantages of Boris Boarman

The major advantage of Boris Boarman compared to other alternatives is that it guides the user through the idea validation and application process by asking sequenced questions based on one of the three scenarios, instead of waiting for the user to come up with the right prompt. Utilizing custom fine-tuned LLM models for each scenario, we minimize the risk of hallucination and maximize the positive outcome from each interaction with the bot.

Team

The team from Whatdahack?! for this project consists of AI engineers, frontend developers, backend developers, project management, design, QA, and marketing personnel.

We are developers and researchers at the forefront of Robotic Process Automation (RPA) solutions, leveraging cutting-edge Artificial Intelligence and Machine Learning tools. With over 14 years of experience, we have developed dozens of web2, SaaS, and API-based development tools. Since 2016, we have been involved in enterprise-level blockchain projects and decentralized applications (dApps) for crypto-native users.

Budget

We have initially estimated the effort to develop:

  • The MVP with the basic functionality of the first scenario at 7500 DOT
  • The full application we envision at about 35000 DOT

The detailed budget breakdown for each milestone will be included in our proposal.

Prototype to Test

We have assembled a very basic prototype, custom GPT, utilizing OpenAI infrastructure and configured it to source knowledge from a few data sources. To see the functionality, you can watch the demo video. To test the prototype, you will need an OpenAI Plus subscription.

Prototype Demo

We have already started work on developing the MVP and will release it as soon as we see positive results on the referendum. Our projected timeline is to have a stable version of the full application by the end of Q4 2024.

As Polkadot technology aims to create a web where data and assets can move seamlessly between multiple blockchains and the AI-bot is designed to provide navigation over the whole crypto universe , we decided to present it first to its community.

You can find the full proposal document here

Up
Comments
No comments here