Inspiration
We are inspired by building a pathway to extend Claude 2 to a billion users.
What it does
Adonais is a privacy preserving API for instant AI monetization, a true win-win. We are demonstrating that by using Claude's operation cost as a floor price for ad auctions we are able to offset the cost of running the LLM by passing the cost to the advertiser.
Our solution:
- Provide easy integration with any conversational app
- Extract topics from the content of the conversation using AI
- Match topics with relevant advertisers
- Capture analytics for developer and advertiser
- Second layer of reinforcement learning
How we built it
We built this on Streamlit by using the globalized reasoning capabilities of the Claude 2 API with langchain and the Google SERP API.
Challenges we ran into
Our biggest challenge was fetching SERP results with production-quality speed which we feel will be easily overcome with a native integration of an enterprise-capable auction API.
Accomplishments that we're proud of
Conceiving a framework that provides the opportunity to make Claude 2 cost-neutral by passing the LLM operation cost to an advertiser in a privacy-preserving manner.
Our privacy-first approach users Claude 2 to extract the topics sent to the ad APIs
Furthermore, we're proud of building a product that uses production APIs to illustrate our vision.
What we learned
We realize that our vision will require extensive collaboration with ad APIs such as Google AdSense, Meta, TikTok, etc but over time we see this model paying for itself while remaining true to Anthropic's mission to 'ensure transformative AI helps people and society flourish'.
What's next for Adonais
- Creative options for developers (personalization + different formats)
- Bidding system for advertisers
- User analytics and account management
Built With
- claude2
- google-serp-api
- langchain
- python
- streamlit
- trubrics
Log in or sign up for Devpost to join the conversation.