badgermole ai

Badgermole ai is an assistant helping the visually impaired navigate life with real-time guidance, intent recognition, and personalized support—bridging accessibility gaps where infrastructure fails.

Comment

🦡 Badgermole AI: An AI-Powered Search Tool for the Visually Impaired

🌟 Inspiration

Living in a rapidly evolving city presents unique challenges, especially for the visually impaired. While technology advances and urban landscapes transform, accessibility often lags behind—whether it’s navigating new infrastructure, identifying stores and restaurants, or finding essential information online.

We created Badgermole AI to bridge the accessibility gap by making information retrieval effortless. Our goal is to empower visually impaired users with an AI-driven assistant that helps them navigate digital spaces just as effectively as they navigate the physical world.

🛠️ How We Built It

Badgermole AI leverages modern web, AI, and accessibility technologies to create a seamless and fast search assistant:

Core Technologies

  • Frontend: Next.js for high-performance rendering.
  • UI Framework: shadcn for beautifully styled components.
  • Runtime & Deployment: Bun for speed, hosted on Vercel using Vercel AI packages.
  • Backend AI & Processing:
    • Groq & Cerebras for ultra-fast inference and AI execution.
    • Perplexity Sonar for real-time AI-powered search.
    • LangChain for intelligent agent toolkits and multi-step AI workflows.
    • Eleven Labs for text-to-speech (TTS), allowing seamless audio output.
    • Whisper AI for speech-to-text (STT), enabling voice interactions and transcription.

🚀 Features

  • AI-Agentic Powered Search: Retrieve relevant information instantly.
  • Intent Classification: AI organizes data for easier navigation.
  • Voice Command Support: Hands-free interaction using Whisper AI for STT.
  • Screen Reader Optimization: Designed specifically for visually impaired users.
  • Real-Time Text-to-Speech: Audio playback of search results using Eleven Labs AI.
  • Multilingual Support: Understands and responds in multiple languages for global accessibility.
  • Money Recognition: Identifies and announces currency for easy transactions.
  • Internet Search Capability: Can look up answers to user queries in real-time.
  • Hazard Detection: Alerts users to obstacles, street crossings, and other safety concerns.

🔥 Challenges We Faced

  1. Optimizing for Speed: Implementing Groq & Cerebras for fast inference while keeping latency low.
  2. Enhancing AI Search: Leveraging Perplexity Sonar to balance accuracy and speed in AI-driven retrieval.
  3. Integrating Multi-Modal AI: Combining Whisper (STT) + Eleven Labs (TTS) + LangChain Agents for smooth AI-assisted voice interactions.

📚 What We Learned

  • How to design AI applications with accessibility in mind.
  • The impact of screen reader-friendly UI/UX in improving user experience.
  • Optimizing AI models using Groq, Cerebras, and LangChain for real-time inference and search.

🏢 Customer Usage

Vision impairment poses an enormous global financial burden, with the annual global cost of productivity estimated to be US$ 411 billion. Many daily tasks that sighted individuals take for granted—navigating a city, making payments, ordering at a restaurant, or accessing educational resources—can be significantly more challenging for the visually impaired.

Badgermole AI is designed to be more than just a digital search assistant—it serves as a real-world accessibility tool to help visually impaired individuals navigate both physical and digital environments.

Use Cases:

  • Navigating the City: AI-assisted location awareness, helping visually impaired users find nearby stores, restaurants, public transport stations, or safe walking routes.
  • Payments at Stores: Providing real-time assistance for handling cash payments or identifying card payment terminals.
  • Ordering at Restaurants: AI-driven menu scanning and voice-guided ordering assistance, allowing users to interact with waitstaff more independently.
  • Students & Researchers in Educational Settings: Assisting with text summarization, AI-powered note-taking, and document organization, making research and learning more accessible.
  • Developers: Quick access to documentation or code snippets using voice commands and screen-reader-friendly interfaces.
  • Businesses & Workplaces: Enhancing knowledge retrieval with an accessibility-first UI, ensuring that employees with visual impairments can access company resources effortlessly.

By combining AI-powered search with real-world accessibility features, Badgermole AI aims to empower visually impaired individuals, allowing them to live more independently and participate fully in education, work, and society. 🚀

🔮 Future Implementations

We plan to expand Badgermole AI with:

  • Advanced Voice Search & AI Chatbots for hands-free interaction.
  • Integration with Cloud Storage (Google Drive, Notion, Dropbox).
  • Real-Time Text-to-Speech Summarization for visually impaired users.
  • Mobile App with Voice Navigation for on-the-go accessibility.

💡 Conclusion

Badgermole AI isn’t just another search tool—it’s a next-generation, AI-powered knowledge assistant built with accessibility in mind. By prioritizing inclusive design, we aim to empower visually impaired users with the freedom to find, organize, and retrieve information effortlessly. 🚀

🔗 Check it out on GitHub: Badgermole AI

Built With

Share this project:

Updates