Idea Detail Analysis
Idea Discovery Interface
ReAct Framework Analysis
Dashboard Overview

About the Project

Inspiration

We were inspired by the challenges organizations face when managing innovation. Many teams collect hundreds of ideas through platforms like ServiceNow Idea Portals but struggle to objectively evaluate which ones deserve resources. The current manual prioritization process is often subjective, time-consuming, and doesn't scale effectively across departments. We saw an opportunity to leverage AI to bring transparency, consistency, and data-driven decision-making to innovation management.

What it does

Service WOW transforms how organizations handle innovation by creating an intelligent layer between idea submission and implementation decisions. Our platform:

Automatically evaluates submitted ideas across multiple dimensions (ROI, effort, strategic alignment, and risk)
Provides natural language explanations for each score, making the AI's reasoning transparent
Visualizes ideas on interactive dashboards that help decision-makers quickly identify high-impact opportunities
Extracts actionable insights from meeting recordings to capture innovation discussions that might otherwise be lost
Enables teams to collaborate through voting and discussion while maintaining a single source of truth

How we built it

We created a full-stack application using a modern tech stack:

Frontend: React with responsive UI components and data visualization libraries
Backend: Cloud Functions to handle the integration with Google's Gemini AI model
Database: Firestore for real-time data storage and retrieval
AI Integration: Structured prompts using the ReAct framework that combine reasoning and action to evaluate ideas systematically
Meeting Intelligence: Speech-to-text processing with sentiment analysis and action item extraction

Challenges we ran into

Designing prompts that would produce consistent, meaningful evaluations across different types of ideas
Balancing the need for comprehensive analysis with response time constraints
Creating visualizations that effectively communicate multiple dimensions of data
Ensuring the system provides useful context for its decisions rather than just numerical scores
Building a responsive interface that works across devices while handling complex data

Accomplishments that we're proud of

Creating an end-to-end system with a full evaluation pipeline in under 10 seconds
Developing an elegant dark/light mode UI that makes complex data accessible
Successfully implementing the ReAct framework to provide step-by-step reasoning for each evaluation
Building a meeting intelligence feature that captures innovation discussions from audio/video files
Designing intuitive visualizations that help users quickly identify the most promising ideas

What we learned

The importance of structured prompting techniques when working with large language models
How to effectively combine quantitative scoring with qualitative reasoning
Techniques for visualizing multi-dimensional data in an accessible way
Methods to process and extract insights from unstructured meeting content
Strategies for creating responsive, real-time applications with Firebase

What's next for Service WOW

Implementing user-adjustable parameters to customize evaluation criteria based on organizational priorities
Adding a feedback loop so the system learns from implementation outcomes
Expanding integrations with project management tools to streamline implementation handoff
Developing more advanced visualization options for portfolio-level analysis
Creating industry-specific evaluation templates to better serve different sectors