DEV Community

aprajita singh
aprajita singh

Posted on

Shiori AI Buddy

Hacktoberfest: Maintainer Spotlight

What I Built & Who It's For

I built Shiori AI Buddy, an open-source AI study companion designed specifically for my classmate and peer who struggled to find organized, low-bandwidth study resources and notes for CBSE Grade 10 prep and language learning.

In resource-constrained environments, students often face unreliable internet access and limited budget for costly AI study tools. Shiori AI Buddy acts as a local-first, privacy-focused study matchmaker that parses notes, creates quick flashcards, and organizes peer-to-peer study exchanges without relying on expensive SaaS APIs.

Why Open Innovation Matters

Open innovation is crucial for making education genuinely accessible. Here is why an open-source approach worked far better than a closed one:

  • Zero Cost & Offline Availability: By leveraging open-weight models (like Gemma), the tool can run locally on budget hardware without needing paid API subscriptions or continuous high-speed internet.
  • Data Privacy: Student notes and personal learning gaps remain entirely on the local device, ensuring complete data ownership.
  • Customization: Open frameworks allow us to fine-tune prompt strategies specifically for Indian CBSE secondary curricula and peer study networks.

How It Works & Technical Details

  • Core Engine: Built around open-source AI agent workflows to index study materials and extract key concepts.
  • Resource Matching: Uses lightweight search and vector mapping to pair students with complementary notes and circular learning resources.
  • Interface: Designed for fast, offline-first access with minimal friction.

Code & Demo

Feedback & Impact

When I shared the prototype with my peer, she was able to instantly search through organized topic summaries offline without worrying about mobile data limits. Her feedback: "It makes finding exact chapter notes so fast without distractions."

Top comments (0)