AutoReportAI

AutoReportAI, built with ADK, turns a Google Sheet into a full report in seconds—handling data ingestion, analysis, visualization, insights, and report generation via intelligent agents.

Comment

Inspiration

In many workplaces and educational environments, data lives in spreadsheets, but extracting insights from that data is still a manual, tedious process. We were inspired to create AutoReportAI to democratize data analysis, allowing anyone—regardless of technical skill—to generate professional-level insights from their data in seconds. With the power of the Agent Development Kit (ADK) and Google Cloud, we wanted to show how multi-agent systems can automate data workflows, making data intelligence accessible, efficient, and smart.

What it Does

AutoReportAI is a multi-agent AI system that takes a Google Sheet URL and automatically performs the following:

  1. Cleans and processes data
  2. Performs basic statistical analysis
  3. Generates smart, human-readable insights
  4. Presents the findings on a user-friendly dashboard

Users simply paste their Google Sheet link and click Generate Report—the agents take care of the rest. There’s also an interactive “How It Works” slider, animated frontend, and a detailed About section.

How We Built It

We built AutoReportAI using: Agent Development Kit (ADK, Python) for designing and orchestrating multiple intelligent agents:

  1. InputAgent: Fetches and parses Google Sheet data.
  2. AnalysisAgent: Computes statistical metrics.
  3. InsightAgent: Crafts contextual insights.
  4. ReportAgent: Orchestrates and presents the findings.

Google Cloud Platform:

  1. Cloud Run for backend deployment
  2. Artifact Registry for Docker images
  3. Google Sheets API with secure key-based access

Frontend:

  1. HTML, CSS, JavaScript
  2. Responsive layout with animations and interactive sliders
  3. Styled using modern design trends from top AI product sites

Challenges We Ran Into

  1. Parsing sheet values: All values came in as strings—this required smart type inference and error handling.
  2. Designing clean inter-agent communication with ADK took several iterations.
  3. Deploying to Cloud Run with optimal memory/runtime settings.
  4. Making a UI that feels modern, sleek, and AI-powered.

Accomplishments That We’re Proud Of

  1. Successfully orchestrated a modular multi-agent system using ADK.
  2. Built an elegant end-to-end cloud-deployed solution with secure data handling.
  3. Created a professional and animated frontend UI that clearly explains the product and enhances trust.
  4. Made data intelligence approachable and usable for non-tech users.

What We Learned

  1. Hands-on expertise with Agent Development Kit (ADK) and how to use agent-based thinking for real-world problems.
  2. Gained deep experience integrating Google Cloud services like Cloud Run, Sheets API, and Artifact Registry.
  3. Understood the importance of user-centric design in building trust in AI tools.
  4. Learned how to design clean modular backend architecture with clear agent roles and orchestration.

What’s Next for AutoReportAI

  1. Visual Dashboards: Auto-generate graphs and trends from data.
  2. LLM-based InsightAgent: Use a large language model to generate more contextual insights and predictions.
  3. Multiple sheet support: Combine and analyze across sheets and tabs.
  4. User history & storage: Let users save and retrieve past reports securely.
  5. OAuth-based Sheet Access: Remove the need for public sheet links and support user auth.

Built With

Share this project:

Updates