Hey guys!
Our team has developed a virtual assistant that can have conversations with users in the NEAR ecosystem. This chatbot uses natural language processing to understand what users are asking and respond accordingly. In this proposal, we will describe the features and capabilities of this technology and how it can be used in the NEAR ecosystem.
The project aims to develop an intelligent virtual assistant (chatbot) that can engage in a full-fledged conversation with a NEAR ecosystem participant, allowing users to submit typical requests for onboarding, submitting proposals/reports for their projects in the NEAR ecosystem.
Currently, the bot is capable of recognizing the context and theme of a user’s question and generating responses based on machine learning. The bot’s internal NLU (Natural Language Understanding) engine allows it to recognize “intents” (requests). To simplify the dialogue navigation, interactive buttons have been implemented to enable users to control the dialogue flow.
The bot uses a knowledge base to generate responses. Currently, the knowledge base used by the bot has been trained on Marketing DAO documents. Additionally, the bot is integrated with ChatGPT to generate responses to more general questions about the NEAR ecosystem and cryptocurrencies in general.
Bot also has own statistics, which reflects the essence of the dialogue, who contacted the bot, topics, and so on. This will enable the integration of additional statistics into MDAO’s work and evaluate the contribution to the ecosystem.
Workflow.
First Stage (June)
-
The bot should be able to recognize the context and topics of a user’s question and generate responses based on machine learning, providing support to NEAR community participants.
-
The bot should operate based on a structured scenario.
-
The bot should use the Marketing DAO Governance Charter document as a knowledge base, considering all updates. The bot should be able to retrain based on the addition of new knowledge.
The bot should be able to:
- Conduct a dialogue in the Telegram messenger
- Create, configure, and modify dialogue scenarios in real-time
- Use reference materials and databases in scenarios
- Test-created or modified scenarios
- Retrain scenarios based on dialogue history
- The bot should have interactive buttons to simplify dialogue navigation.
- The bot should also provide dialogue statistics and analytics, highlighting the most frequent user requests.
Second Stage (Extended Functionality) (July)
- Expand the bot’s knowledge base to include all DAOs in the Near Protocol ecosystem
- Expand the bot’s functionality to enable interaction with NEAR systems (transmitting and receiving various types of data between the bot and the user)
- Integrate the bot with the NEAR platform, including wallet authorization through the web app and Telegram
- Create a widget on the NEAR website (https://near.org/)
Budget
The budget is divided into stages.
| Stage | Budget |
|---|---|
| June | $6 000 |
| July | $6 000 |
| Total | $12 000 |
Proposer: lolson.near
KYC passed.