Inspiration

As hybrid and remote teams rely more on virtual meetings, understanding how time is spent in those meetings becomes essential. Yet, most organizations lack a clear view of how meetings affect productivity, collaboration, or workload balance. MeetDash was created to help organizations easily centralize, analyze, and interpret their Google Meet activity data — empowering teams with actionable insights for better time management and efficiency.

What it does

MeetDash is an internal analytics tool that helps organizations sync and analyze Google Meet activity seamlessly. Automatically syncs meeting metadata (duration, participants, timestamps) from Google Meet via a custom Fivetran connector

Runs automated transformations and analytics in the backend to compute: Total and average meeting counts Meeting duration and overlap trends Recurring meeting patterns Department or user-level meeting load

Uses Gemini API to generate summaries of meeting insights and trends. Presents results through a clean analytics dashboard for internal stakeholders.

How we built it

Built a new Fivetran connector using the Connector SDK to extract structured meeting data from Google Meet APIs. Synced data into Google CloudSQL using Fivetran’s automated pipelines. Applied transformation logic to compute key meeting metrics (frequency, duration, overlaps). Integrated the Gemini API for summarizing metrics and generating AI-driven insights. Visualized insights through an internal dashboard.

Challenges we ran into

Limited direct APIs for Google Meet usage data, requiring creative extraction and mapping. We could not use Realtime data from meet transcript due to workspace restriction from google. Designing schema transformations for clean incremental updates. Integrating Gemini summarization into backend workflows efficiently.

Accomplishments that we're proud of

Developed a new Fivetran connector for Google Meet — a data source not natively supported. Built a fully automated analytics pipeline from ingestion to insight generation. Used Gemini API to make analytics understandable through concise natural language summaries. Created an internal dashboard that gives managers and ops teams real visibility into meeting efficiency.

What we learned

How to extend Fivetran with a custom data connector for unsupported APIs. The power of combining data automation (Fivetran + CloudSQL) with AI summarization (Gemini). How small internal datasets, when centralized, can drive operational improvement. Effective schema design and incremental loading for time-series meeting data.

🚀 What’s next for MeetDash Integrate additional collaboration tools like Google Chat or Slack to correlate discussion activity with meeting load. Add predictive analytics (e.g., forecasting meeting fatigue trends). Enable cross-department benchmarking for operational planning. Expand the dashboard into a lightweight SaaS module for enterprise clients. Add a chatbot for org owners to interact using natural language.

Built With

• fivetran
• gemini
• google-cloud
• googleaicloud
• javascript
• meet
• python