RxQuery.AI

No more googling pills, your AI Pharmacist is here for smarter, faster and reliable medicine information!

Comment

Inspiration ✨

RxQuery.AI was inspired by the need for an intelligent, AI-powered drug information finder and medicine assistant, essentially an "AI Pharmacist." The goal was to provide instant, reliable drug consultation through an intuitive interface, replacing the need for "Dr. Internet" for drug information.

What it does 🎯

RxQuery.AI is an intelligent, AI-powered drug information finder and medicine assistant that offers instant, reliable drug consultation. It's deeply integrated with MindsDB's Knowledge Bases, Agents, AI Tables, and Evaluation Tools to provide reliable drug insights powered by semantic search and LLM Agents.

⚠️ Note: RxQuery.AI is not a substitute for professional medical advice. Please consult actual doctors, thanks :)

✨ Key Features

General Purpose Assistant: RxAssistant (via AI Table) AI Agents:

  • 🔍 Drug Classification - Instantly classify medications (Antibiotic, Analgesic, etc.)
  • 💊 Smart Recommendations - Get personalized drug suggestions based on symptoms
  • ⚠️ Side Effects Checker - Comprehensive side effects analysis
  • 🛡️ Allergy-Safe Search - Find safe alternatives for patients with allergies Command based input:
  • ⌨️ Command-based Interaction - Use simple slash commands like /classify, /recommend, etc.
  • 🎙️ Voice Interaction - Talk to RxQuery for hands-free health queries
  • 🎨 Beautiful UI/UX - Chat-based interface with real-time updates

How we built it

  • Frontend: Next.js 14, TypeScript, Tailwind CSS, Framer Motion, Shadcn, MVPBlocks
  • Backend: FastAPI, Python, Pydantic

  • 🧠 MindsDB

    • CREATE KNOWLEDGE_BASE, INSERT INTO, CREATE INDEX(chromadb)
    • CHAINED Multi AGENTS using CREATE AGENT for each feature (/classify, /recommend, /side-effects, etc.)
    • EVALUATE KNOWLEDGE_BASE with Groq for document scoring
    • CREATE JOB to ingest drug data periodically
    • SELECT ... WHERE content LIKE in semanticss!
    • metadata_columns to enable hybrid semantic + SQL filtering
    • CREATE MODEL rx_assistant with OpenAI for reasoning and classification
    • KB_EVALUATE: Groq LLM, AI TABLES: OpenAI, AGENTS: OpenAI, Ollama (experimental, model removed)
    • 🧪 Editor: Our MindsDB SQL Editor code is included as reference for building/debugging Agents and KB queries.

Challenges we ran into

Building RxQuery.AI definitely came with its share of hurdles. One of the main challenges was integrating the various MindsDB components seamlessly. While MindsDB offers powerful features like Knowledge Bases, Agents, and AI Tables, getting them all to communicate effectively and consistently deliver accurate, nuanced drug information required a lot of fine-tuning. I had to spent a good amount of time experimenting with different prompt engineering techniques for our agents to ensure they could handle a wide range of medical queries, from simple classifications to complex side-effect analyses.

Another significant challenge was curating and structuring the drug data for our Knowledge Bases. Ensuring the data was comprehensive, up-to-date, and in a format that the AI could easily interpret for semantic search was a continuous process.

Accomplishments that we're proud of

I'm incredibly proud of several key accomplishments with RxQuery.AI. Firstly, successfully creating an intelligent, AI-powered drug information system that feels like a real-time pharmacist. The ability to provide instant, reliable drug consultations through an intuitive chat interface, powered by AI knowledge, is something I believe fills a critical gap.

I'm particularly proud of the deep integration with MindsDB. Leveraging their Knowledge Bases, various chained AI Agents (like /classify, /recommend, /side-effects, and /allergy), and AI Tables for general queries really showcased the platform's power. It was a huge win to abstract complex AI models into these user-friendly components. Seeing the system accurately classify medications, recommend drugs based on symptoms, and check for allergies with high precision is a big win.

What we learned

This project was a massive learning experience, especially in the realm of AI-driven applications for specialized domains. I gained invaluable insights into building robust AI agents and knowledge bases within the MindsDB ecosystem. Understanding how to effectively CREATE KNOWLEDGE_BASE, INSERT INTO, and CREATE INDEX for efficient semantic search was fundamental. I also deepened my knowledge of prompt engineering and the nuances of training and fine-tuning LLMs for specific tasks, ensuring our agents provided medically relevant and safe information.

What's next for RxQuery.AI

The current version could be developed further to provide intelligent responses and to be more accessible to people. I'm looking to integrate Prescription management and Multi-language support in the upcoming versions.

Built With

Share this project:

Updates