ByeFrust AI

Inspiration

In today’s digital marketplace, consumers are overwhelmed by biased reviews, affiliate marketing, and complex technical jargon. This often leads to confusion and poor purchasing decisions. We were inspired to create a trustworthy AI companion that simplifies decision-making and empowers users with transparent, unbiased recommendations.

What it does

ByeFrust AI is an intelligent recommendation system that interacts with users through a dynamic interview to understand their needs. It applies hard filters such as budget and must-have features, ranks products using a deterministic weighted scoring engine, and presents the top 3–5 options with personalized pros and cons. The system ensures transparency and eliminates commercial bias.

How we built it

The prototype was developed using Streamlit for the conversational user interface and Python for backend logic. A custom weighted scoring engine ranks products based on user-defined priorities. Google Gemini API is integrated to generate natural-language explanations and trade-off analyses. Product data is currently stored in CSV files, with plans to migrate to PostgreSQL and external APIs for production.

Challenges we ran into

Designing an unbiased and deterministic scoring mechanism.
Creating adaptive interview logic that responds dynamically to user inputs.
Integrating LLM-generated explanations without influencing the ranking process.
Managing limited and structured product datasets for the prototype.

Accomplishments that we're proud of

Built a fully functional prototype within a short ideathon timeline.
Developed a transparent scoring engine independent of LLM bias.
Created an adaptive interview system for personalized recommendations.
Delivered explainable AI with clear pros, cons, and confidence scores.

What we learned

The importance of explainability and transparency in AI-driven decisions.
Effective integration of LLMs as supportive tools rather than decision-makers.
Designing user-centric systems that balance automation with control.
Rapid prototyping and teamwork under time constraints.

What's next for ByeFrust AI

Integration with real-time product APIs for up-to-date data.
Expansion into multiple product categories such as wearables and home appliances.
Development of a voice-first mobile application.
Implementation of user feedback loops for continuous learning.
Migration to scalable cloud infrastructure.