VibeCheck is an advanced AI-powered application designed to solve "Analysis Paralysis" in consumer decision-making. By synthesizing real-time web reviews into a concise, actionable sentiment report, VibeCheck bridges the gap between fragmented consumer feedback and objective decision-making.
- Real-Time RAG (Retrieval-Augmented Generation): Unlike static AI models, VibeCheck fetches the latest review data from across the web using the Serper API, ensuring insights are always current.
- Aspect-Based Sentiment Analysis (ABSA): Automatically extracts specific Pros and Cons, distilling noisy text into structured, high-value insights.
- Signal Strength Quantification: Provides transparency by reporting the amount of data analyzed, allowing users to gauge the reliability of the verdict.
- Advanced Sentiment Engine: Powered by Google Gemini 3 Flash, the system performs semantic reasoning to detect sarcasm, context, and nuance, moving beyond simple keyword counting.
- Zero-Latency Pipeline: Built on a browserless, REST-based API architecture for rapid performance.
- Frontend & Backend: Streamlit (Python)
- Intelligence Engine: Google Gemini 3 Flash
- Data Retrieval: Serper.dev API
- Visualization: Plotly
- Deployment: Streamlit Cloud
- Retrieval: User input triggers the Serper API to scrape top-tier review snippets from diverse sources (Reddit, Amazon, Tech Blogs).
- Augmentation: Retrieved snippets are aggregated into a structured prompt, providing the AI with the necessary context ("grounding").
- Generation: The Gemini 3 Flash model parses this context to output a structured JSON verdict (Sentiment Score, Vibe Summary, Pros, and Cons).
- Presentation: Streamlit visualizes the sentiment using dynamic gauge charts and responsive UI components.
- Clone the repository:
git clone https://github.com/yourusername/vibecheck.git cd vibecheck - Install dependencies:
pip install -r requirements.txt
- Configure Environment Variables:
Create a
.envfile and add your API keys:GEMINI_API_KEY=your_gemini_key_here SERPER_API_KEY=your_serper_key_here - Run the application:
streamlit run app.py
- Hallucinations: Mitigated by strict RAG grounding and JSON-only structural constraints.
- API Limits: Handled via exponential backoff retry logic and intelligent model fallback.
- Data Bias: Addressed by pulling from multi-platform sources to achieve a consensus-driven sentiment.
Developed as an engineering project exploring the frontiers of Natural Language Understanding (NLU) and Dynamic Information Retrieval. VibeCheck demonstrates the efficiency of modern LLMs in high-speed, real-world data processing tasks.