apneduniya commented 1 month ago

Varta

Submitter Information

Name: Adarsh Gupta, Email: thatsmeadarshgupta@gmail.com

Reward Address

0x17Ad46C3C8AF35c79995014173Bf263DD14d0A22

Project Plan

My goal is to create a News Feed Analyser app which provides comprehensive summaries of the latest news from many news outlets based on user interests, leveraging open-source LLMs, ScrapeGraph AI, Spheron and tools like LangChain ️.

My upcoming goal is to:

Currently, it supports 14 news outlets and 60 interests which I want to extend to 30-40 news outlets which is popular among the news/article readers.
Currently, it uses Llama3 8B model and I want to add support for multiple open source models like Llama3 70B, Mixtral-8 7B, Llama-3.1 8B so that the users can choose freely.
User can chat with the news, so if they have some doubts or unclear about somethings then the Chatbot can help them.
Adding a pricing/subscription model and will charge users for more frequent newsletters on email, using models other than default (Llama3 8B), rate limits in Chatbot.

Repository

https://github.com/apneduniya/varta

Demo

Video Demo: https://www.youtube.com/watch?v=OL2nFczDWso

Setup Instructions

https://github.com/apneduniya/varta?tab=readme-ov-file#-steps-to-run

Spheron Infrastructure Usage

Spheron played the crucial role in the deployment of this project. By leveraging Spheron's robust and scalable deployment platform, we ensured that both our server and open-source AI model is accessible and performant on mobile app.

Checklist

[x] Added appropriate bounty label
[x] Provided all required information
[x] Ensured demo links are working
[x] Verified setup instructions are clear and complete

rekpero commented 3 weeks ago

@apneduniya, can you mention what open-source tool, library, and tech stack has been used on your project? also mentioned all the functionality that you have implemented.

apneduniya commented 3 weeks ago

@rekpero I have mentioned the features in the README.md of the repository, but also mentioning here:

🗞️ Aggregate news from 14+ sources.
🔍 Analyze content from 60+ interests choosed by user.
📝 Concise and comprehensive summaries of the latest news.
📧 Email subscriptions for newsletter with customizable frequency (daily, weekly and monthly).
🔖 Bookmark news for later read.
👤 User Authentication (Login/Register).

The tech stack I have used are: React Native (Expo), FastAPI, Langchain, MongoDB, ScrapeGraph AI, Ollama Llama 8B, SMTP, Spheron and BS4

apneduniya commented 2 weeks ago

@izrake https://www.youtube.com/watch?v=j0LiRiz2iXg Video on how to deploy the backend on Spheron and add the backend link in app to start it :)

izrake commented 2 weeks ago

Hey @apneduniya, this looks quite good.

You have passed all the parameters of the bounty & happy to approve the 1500 USD for this work.

We will release the bounty amount on this address 0x17Ad46C3C8AF35c79995014173Bf263DD14d0A22

But before that, we will send you an email to confirm some KYC details. Please look out for it.

apneduniya commented 2 weeks ago

Thanks a lot sir, super excited to start my work on upcoming goals 💪

izrake commented 4 days ago

Funds are released https://etherscan.io/tx/0x81c987e2b8c0828f5563402d0d35a911551d39d4a77e29b0046d9a690be604b6

I'm closing this issue & please make sure you don't make your repo private, which you have attached here.

spheronFdn / sos-ai-bounty

Varta - A News Feed Analyser APP #10