Bright Day News Application

Application landing page screenshot
Application screenshot
Application screenshot
Bright Day News Application Architecture Diagram

BrightDay Application – Project Summary and Application Architecture

Project Overview

With so much bad news everywhere, I felt the need to create something that reminds us there's still a lot of good in the world. This app is my way of sharing uplifting stories that can brighten someone’s day and bring a bit more balance to how we see things.

BrightDay is a web application that uses Agent Development Kit (ADK) to deliver curated positive news and reflective content to users in an engaging, modern interface. The primary goal of the project is to automate the orchestration of specialized AI agents—via a coordinating root agent—to gather uplifting news stories from across the internet, summarize them, and generate reflective commentary. The application emphasizes transparency of process through workflow visualization.

Key Features and Functionality • Multi Agent Orchestration: Uses ADK Python SDK to create multi agent system. Root agent orchestrates two sub-agents: SearchAgent and ReflectAgent. Root agent is utilizing Google Gemini model gemini-2.0-flash. • Good News Aggregation: Uses a ADK SearchAgent to perform web searches for positive, uplifting news stories for a user-specified date. Search Agent is utilizing Google Gemini model gemini-2.0-flash, and Google search tool. • Automated Summarization and Commentary: Employs a ReflectAgent to generate a thoughtful, emotionally resonant commentary based on the gathered news. Reflective Agent is utilizing Google Gemini model gemini-2.0-flash. • Workflow Visualization: Displays a real-time step-by-step execution of the AI agents for transparency and debugging. • Persistent Session Storage: Stores user prompts, news articles, and reflections in a relational PostgreSQL database. • User-Friendly Interface: Responsive UI • Session History: Users can browse past sessions and revisit results.

Learnings and Reflections Technical Learnings • ADK is a robust and powerful AI Agent system for orchestrating AI agents • Asynchronous agent management was essential to ensure responsiveness; leveraging async/await improved performance and reduced UI blocking. • Workflow visualization added significant debugging and transparency value, but required careful synchronization with backend state. • Challenges with parsing inconsistent responses—solved through layered JSON repair strategies.

Conclusion

BrightDay News demonstrates how AI agents, when coordinated through a modular, well-architected application, can provide meaningful and emotionally uplifting experiences. By combining modern web technologies, a robust backend, and transparent AI workflows, the app provides a unique take on automated content curation with a human-centered twist.