Inspiration

Managing personal administration has become increasingly difficult as people receive more digital information than ever before. Important details such as subscription renewals, bills, invoices, warranties, and payment deadlines are often buried inside emails and documents.

I was inspired to build Tendly because many people struggle to keep track of these small but important responsibilities. Missing a cancellation deadline or forgetting a recurring payment can lead to unnecessary costs and stress.

The name Tendly comes from the word "tend" because Tendly helps users tend to the things that matter in their lives.

My goal was to create an AI-powered assistant that does more than store information. I wanted to build a system that understands users' administrative tasks and helps them take action at the right time.


What it does

Tendly is an AI-powered personal administration assistant that transforms emails and documents into organized records, deadlines, and actionable insights.

Users can securely connect their Gmail account, allowing Tendly to analyze emails and identify important information such as:

  • Subscription services
  • Payment amounts
  • Renewal dates
  • Cancellation deadlines
  • Important confirmations

Using AI, Tendly extracts relevant details and organizes them into a clean dashboard where users can easily view their responsibilities.

The platform also provides:

  • Gmail synchronization to automatically discover important information
  • AI-powered extraction from emails and uploaded documents
  • Review workflow where users can verify AI-generated information before saving
  • Deadline reminders to help users take action before important dates
  • Analytics dashboard to visualize spending patterns and subscriptions
  • Document upload support for invoices, screenshots, and confirmations

Tendly helps users reduce digital clutter and stay in control of their everyday responsibilities.


How I built it

I built Tendly as a full-stack web application combining modern web technologies with AI services.

The frontend was developed using React, TypeScript, Vite, and Tailwind CSS to create an intuitive dashboard experience where users can view subscriptions, deadlines, and analytics.

The backend was built using Node.js, Express.js, and TypeScript to handle authentication, API communication, data processing, and application logic.

For user authentication and email integration, I implemented:

  • Google OAuth 2.0 authentication
  • Gmail API integration
  • Secure read-only Gmail access

For AI processing, I integrated Gemini AI to analyze emails and documents. The AI pipeline extracts structured information such as vendors, amounts, dates, categories, and transaction details.

The extracted results are validated and placed into a review workflow before being stored, ensuring users maintain control over AI-generated information.

The application uses Prisma ORM with a database layer to manage user records, extracted information, and application data.


Challenges I ran into

One of the biggest challenges was handling the variety of formats found in emails and documents. Different companies structure their invoices, subscription emails, and confirmations differently, making consistent information extraction difficult.

Another challenge was improving AI reliability. While AI can extract information quickly, incorrect assumptions can affect user trust. To solve this, I implemented a review process where users can verify and edit extracted information before saving it.

Designing the dashboard was also challenging because I needed to present a large amount of information without overwhelming users. I focused on highlighting important actions, such as upcoming deadlines and subscriptions requiring attention.

Integrating multiple systems together, including Gmail authentication, email processing, AI extraction, reminders, and analytics, required careful planning to ensure a smooth user experience.


What I learned

Building Tendly taught me how AI can be applied to solve practical everyday problems.

I learned that creating a useful AI application requires more than just connecting an AI model. A successful AI system needs:

  • Reliable data processing
  • User trust and transparency
  • Clear user experience design
  • Validation of AI-generated results

I also learned how to integrate external APIs, design full-stack architectures, and build workflows that transform unstructured information into meaningful insights.

Most importantly, I learned that AI can reduce the mental burden of managing everyday tasks and help people focus on what truly matters.


Future Improvements

Moving forward, I plan to expand Tendly with more intelligent features, including:

  • Calendar integration for automatic scheduling
  • More advanced AI recommendations
  • Automatic subscription cancellation assistance
  • Mobile application support
  • Receipt and PDF scanning
  • Personalized financial insights

The long-term vision for Tendly is to become a proactive personal AI assistant that helps users manage their digital lives effortlessly.

Built With

Share this project:

Updates