Discover how AI Research Assistant is transforming the way professionals conduct research. This powerful web application combines voice input, real-time LLM responses, and intelligent conversation management to create the ultimate research tool.
In today’s information-driven world, conducting efficient research is critical. Whether you’re a student, researcher, content creator, or professional, spending time searching for answers and organizing information is time-consuming. What if you had a personal AI research assistant available instantly?
Introducing AI Research Assistant—a responsive web application that acts as your intelligent research companion, accepting both voice and text input, streaming responses from powerful language models, and maintaining your conversation history with intelligent topic recognition.
What Makes It Unique?
Voice and Text Input: Speak your questions directly using voice-to-text capabilities, or type if you prefer. The app automatically debounces input and sends it to your LLM after a short delay.
Multiple LLM Backends: Choose your preferred AI provider:
- Local Ollama for privacy and control
- OpenAI for powerful cloud-based models
- Custom endpoints for specialized needs
Real-Time Streaming Responses: Watch as responses appear word-by-word in real-time, creating a more engaging and responsive experience.
Full Markdown Support: All responses render with complete markdown formatting including headers, lists, code blocks, and more for better readability.
Conversational Memory: The application remembers your recent conversation context, allowing you to ask follow-up questions naturally across sessions.
Quick Response Feature: Get context-aware one-paragraph summaries of your query topics before the full response, using a lightweight model.
Key Features
✓ Voice-to-text input with automatic submission
✓ Real-time streaming LLM responses
✓ Three-column responsive layout
✓ Topic navigation sidebar
✓ Conversation history with timeline
✓ Word cloud visualization of topics
✓ Query queue system for continuous conversation
✓ Generate HTML reports of conversations
✓ Dig deeper research via Perplexica integration
✓ Enable/disable auto-submission toggle
✓ Bootstrap 5 responsive design
✓ Full markdown rendering
✓ Optional local Whisper for transcription
Responsive Design
The application features a modern three-column layout:
- Left Column: Topic navigation widget showing your conversation topics
- Center Column: Main input area and chat interface
- Right Column: Quick response section and status updates
The layout automatically adapts to different screen sizes, making it work perfectly on desktop, tablet, and mobile devices.
Perfect for Voice-to-Text Workflow
The application is specifically designed to work with voice-to-text utilities like Windows 11 Voice Assistance. Simply:
- Start recording your question
- The app automatically submits after your speech ends
- Receive your response while continuing to take notes
No keyboard clicking required—just speak and read responses.
Advanced Features
Conversation Reports: Generate comprehensive downloadable HTML reports of your conversations, including AI-generated summaries, keywords, and full transcripts.
Topic Analysis: The history page features dynamic word clouds that visualize the most frequent and recent topics in your conversation.
Streaming Responses: LLM responses stream to you in real-time, improving perceived performance and creating a more natural conversational experience.
Persistent Storage: Browser storage maintains your conversation history and context even after closing the application.
Installation and Configuration
- Clone the repository
- Copy .env.example to .env and configure your LLM provider
- Run php -S localhost:8000 from the project root
- Open http://localhost:8000 in your browser
Environment variables include:
- PROVIDER: Choose ‘ollama’ or ‘openai’
- OLLAMA_URL and OLLAMA_MODEL: For Ollama setup
- OPENAI_API_KEY: For OpenAI setup
- DEBOUNCE_MS: Delay before auto-submission
- PERPLEXICA_URL: Optional Perplexica integration
Use Cases
- Students: Research papers and assignments
- Content Creators: Generating ideas and outlines
- Professionals: Quick reference and problem-solving
- Researchers: Deep dive investigations with conversation history
- Anyone seeking quick answers: Without distractions
Conclusion
AI Research Assistant represents a paradigm shift in how we interact with AI for research and learning. By combining voice input, real-time responses, intelligent conversation management, and a beautiful responsive interface, it creates an unmatched research experience.
Whether you’re a power user seeking to optimize your research workflow or someone exploring AI-assisted research, this application provides the tools you need.
GitHub Repository: https://github.com/SPHillips1337/AI_research_assistant