Aintrest

Teachers are overworked, and the students are under looked in India. We provide A 24/7 available personalised teacher to the students and A 24/7 available assistant to teachers and so much more.

Comment

Inspiration

There is the vast gap in education, especially in tier 2 and tier 3 cities and towns, and that's where most Indians live. About 70% of the country resides there, but many lack good teachers. Surprisingly, almost everyone has smartphones. We want to change things. We aims to spark curiosity among kids in these neglected areas. It's not just about better grades; it's about helping teachers teach better and keeping parents in the loop about their child's education. This startup is our way of making a real impact and bridging the educational gap for kids across the nation, while being the most bang for the buck investment in education space.

What it does

AI-Powered Text Interaction: Dive into Indian school course books. Utilise AI assistants to ask questions, request summaries, explanations, self-assessments, and much more.

Writing Bots: Learn how to write Essays, Articles, Diaries and Notices with our Bots designed to help children of all skill levels.

Voice Interaction: Every message is equipped with a text-to-speech button, immersive audio experiences. Additionally, enjoy the convenience of speech-to-text, allowing seamless conversion of spoken words into written text.

Historical Characters: Talk to historical characters, as they were your mates or teacher or enemy, create characters and talk to them as if they were talking to you in real time.

How we built it

Tech Stack Overview Our application, built with Next.js 13, React-Query, and TypeScript, delivers a seamless user experience. Tailwind CSS enhances styling, while Prisma ORM connects to PostgreSQL for efficient data handling.

  • trpc (Typed RPC): Ensures type-safe APIs, providing reliability in communication.
  • PostgreSQL: Our reliable and scalable database solution.
  • Azure Services: VMs, Azure SQL Data, BlobStore, and Text-to-Speech (TTS).
  • LLMs: Power the Assistant API, Chat Completion API, and Embeddings API .
  • Pinecone: Optimises vector data storage and retrieval.
  • Langchain: Streamlines development with abstractions.

Challenges we ran into

Never having used Next.js, TypeScript, React-Query, trpc, and Azure before, my journey with these technologies was quite a ride. There were some issues with setting up web-sockets in nextjs and trpc Most of my time went into reading docs, debugging type errors, and streaming texts or audios. Handling PDFs seemed like a big task at the start. Prompt engineering more like banging head against a wall.

Accomplishments that we're proud of

Learning new tech and implementing it while our minds were cluttered with what to build first. Many aspects need improvement in the built features, but we're committed to fixing them.

What we learned

While building the app, I learned about trpc, Next.js, TypeScript, React-Query, web-sockets and also got into Azure and LLMs.

What's next for Aintrest

We plan launch the MVP before April. We're set to partner with a small school, piloting our product to iterate and find PMF.

Progress

I started this project on 4th feb, basically the whole website was made in the period of the hackathon

Built With

Share this project:

Updates