Ads.ai

Inspiration

Businesses want to integrate AI chatbot features into their products at an extra cost. We can help them recover that cost by seamlessly integrating ads into chat responses. This advertising technique works effectively because it captures the user’s current needs and matches them with specific ads.

What it does

It recommends advertisements based on user conversations and encourages users to engage with the advertised products. It also logs ad metadata to a database for data persistence and traffic tracking, and sends reminder emails about the products based on chat history and the user’s email when logged in.

How we built it

We created a sample chatbot that uses the Lava API to route requests to a specific LLM model. Then we built a Claude MCP server that connects to our SQLite database containing partner advertisements, which the Claude AI model uses to select specific ads based on the user’s prompt. We also used Composio AI agents to log ad metadata to Google Sheets and send reminder emails through Gmail once the user clicks an advertisement link.

Challenges we ran into

Given the short timeline of a hackathon, we couldn’t create a fully detailed automation workflow that persists complete ad metadata and tracks user behavior. We focused on developing a simple working prototype to demonstrate our idea.

Accomplishments that we're proud of

We believe this idea is very practical since many businesses and startups have AI chatbots but struggle to monetize their products. This project helps them earn extra revenue while providing quality service to their clients. We also managed to use as few tools as possible to reduce extra costs.

What we learned

We learned how to build AI workflow automation and an MCP server that connects to our database.

What's next for Ads.ai

We plan to refine our automation workflow to cover more use cases and user behaviors. We will add more AI agents to track user interactions with the chatbot and recommend better ads. We also plan to expand beyond AI chatbots to other types of AI products that involve user input, allowing ads to be integrated into any AI experience.