IdeaGenie

Landing Page
Features on landing page
Dashboard
Allows business user to set business intent
Allows business user to set business intent with constraints
Calculating the the top 3 ideas
Final Results

Inspiration

Identifying the Problem: Organizations struggle to sift through overwhelming numbers of ideas. Leveraging AI: Using a ReAct (Reasoning + Action) framework with an LLM to inject intelligence into idea prioritization. User-Centric Design: Emphasis on a clear UI that not only ranks ideas but also explains the reasoning behind each decision.

What it does

Idea Evaluation: Analyzes ideas based on estimated implementation effort and potential ROI. Feedback Loop: Uses historical company data to improve recommendations over time. Custom Constraints: Incorporates business constraints to fine-tune the evaluation. Explainability: Provides detailed reasoning for the ranked ideas. Interactive Dashboard: Displays an engaging user interface for exploring ranked ideas and insights.

How we built it

Backend Development:

Developed a Flask API that exposes the evaluation service. Created the core LLM evaluation module (p8V3.py) to process idea inputs and historical data. Integrated a feedback loop to continuously refine recommendations.

Frontend Development:

Built a dynamic React-based dashboard. Designed an intuitive interface to display evaluation results and reasoning. Enabled interactive elements for adjusting business constraints and evaluation parameters. Collaboration: Combined expertise from product, engineering, and data science teams to create a comprehensive solution.

Challenges we ran into

Data Integration: Merging historical performance data with new idea submissions. LLM Tuning: Balancing automated reasoning with human interpretability. UI/UX Design: Ensuring that the decision-making process is transparent and accessible. Scalability: Creating a solution that scales with increasing idea submissions and data.

Accomplishments that we're proud of

Successfully integrating a state-of-the-art LLM to simulate ReAct-style reasoning. Developing an interactive dashboard that clearly explains the ranking decisions. Building a feedback loop that enhances the model's evaluation over time. Delivering a solution that is both technically advanced and user-friendly.

What we learned

Interdisciplinary Collaboration: The value of merging insights from various domains. Explainability in AI: How critical it is to communicate the reasoning behind automated decisions. Iterative Improvement: The importance of continuous feedback in refining both backend logic and UI design.

What's next for IdeaGenie

Enhanced Customization: Allow users to save custom configurations and weight settings. Expanded Data Integration: Incorporate additional data sources to further refine evaluations. Real-Time Feedback: Develop real-time dashboards to monitor implemented ideas. User Testing: Gather further user feedback to drive future improvements.