Inspiration
We've worked with talent agencies in Mexico that need not only a CRM, but a powerful tool for connecting their talent with casting directors. We are using AI Agents to bring this to life.
What it does
It allows talent agencies to receive talent profiles from users and provide casting directors and agency owners an easy way to create casting lists.
How we built it
We used Agent Strands and Bedrock to power the AI Agent, we used RAG for the semantic retrieval for the actors profiles and we developed all application layer and scalable infrastructure for real production workflows.
Challenges we ran into
We could've used more AWS managed services like AgentCore to solve a few of the paint points of having production-ready multi-tenant customer-facing applications.
Accomplishments that we're proud of
Being able to define all our infra as code, create useful agent tool calls and retrieval systems for the real user case of the clients.
What we learned
That AWS is probably the best way to start developing production-grade AI Agents, we want to balance the AI Agent development and App Infra for better agile and productive development workflows.
What's next for Talent Agent
We are taking this to production with local talent agencies in Mexico!
Repository
Built With
- agent
- amazon-web-services
- bedrock
- cognito
- dynamodb
- fastapi
- lambda
- postgresql
- pulumi
- rag
- rds
- react
- s3
- strands
- strandsagents
Log in or sign up for Devpost to join the conversation.