Eureka Slack Bot

Eureka is a Slack bot that uses LangChain and OpenAI to have smart conversations. It learns from personal information to talk in a way that feels personal and makes sense. It can also search the internet to give better answers and help more effectively.

Demo of Current Progress

. . . .

Key Features Developed So Far

• ✅ The bot only responds when mentioned in the Slack channel
• ✅ Utilizes LangChain and OpenAI to have intelligent conversations
• ✅ Eureka learns from specific personal data to provide personalized and relevant interactions
• ✅ Capable of fetching online information to supplement responses
• ❌ Instead of fetching data from a local folder, Eureka will integrate with a vector database for improved performance and scalability
• ❌ Ability to handle and read schedules from my personal calendar
• ❌ Scheduling feature to send reminders or automated messages at specific times
• ❌ Integration with external APIs to perform specific tasks, such as fetching news, etc
• ❌ Multimedia responses, including sending relevant images, videos, or files

Technologies Used

• LangChain: Framework for enabling intelligent conversations and natural language processing.
• OpenAI: AI platform used for generating human-like text and understanding language.
• Vector Database Integration: Utilizes a vector database for efficient data storage and retrieval, enhancing performance and scalability.
• Slack SDK: Software development kit for building Slack integrations and bots.
• Python: Programming language used for implementing Eureka's functionality.
• Other Libraries and Tools: Additional dependencies specified in the requirements.txt file to support various features and functionalities of Eureka.

Running the Bot

To run the bot, execute the main.py file:

python src/main.py

Want to Contribute?

Please read our CODE_OF_CONDUCT.md for more details.

Contributors

• @mkafonso

---

. . . . . . .

TESTES QUE FIZ RECENTEMENTE

. . . . . . .

• É tranquilo gerar esses embeddings com a OpenAI

• Depois de pegar o array dos números, precisamos salvar num banco de dados vetorial. Escolhi o SingleStore: https://www.singlestore.com/

• Cria um workspace, database e depois uma tabela

• Cadastrar os dados

• Exagerei e quis subir todo script do filme Shrek 😅 e o modelo reclamou.