Inspiration
The inspiration came from the immediate need to process sentiment from users across the service stack. There are existing solutions on the market but none of them provide the same low cost and scalability options. The simple UI that is part of this package provides an easy way to monitor, configure and observe the solution in action.
What it does
InsightModAI Agent is a serverless, autonomous AI agent that processes customer feedback, analyzes sentiment, identifies themes, and generates actionable insights for business operations. Built entirely on AWS with Amazon Bedrock AgentCore Runtime, it automates customer insights workflows with enterprise-grade reliability.
Autonomous Processing: Automatically processes customer feedback using advanced AI models for sentiment analysis Real-time Insights: Extracts themes, validates sentiment, and generates reports in real-time CRM Integration: Optional integration with Salesforce, HubSpot, and other CRM systems Visual Dashboard: React-based admin interface for monitoring and configuration Enterprise Security: AWS Cognito authentication with least-privilege IAM roles Monitoring: Comprehensive CloudWatch monitoring and alerting Cost Optimized: Serverless architecture with pay-per-use pricing
How we built it
I have utilized serverless services and modular design to achieve the most modern and performant solution possible. Using Bedrock AgentCore runtime proved to be the best option for this solution to bring it to life. The solution is a self contained fully automated package that can be utilized immediately.
- Amazon Bedrock AgentCore Runtime - Containerized Strands agent for AI processing
- AWS Lambda Functions - Serverless compute for feedback ingestion and processing
- Amazon DynamoDB - NoSQL database for feedback storage and configuration
- Amazon S3 - Object storage for processed insights, reports, and static assets
- Amazon API Gateway - REST API for external integrations
- AWS Amplify - React dashboard hosting with Cognito authentication
- Amazon CloudWatch - Monitoring, logging, and alerting ## Challenges we ran into The biggest challenge during the project was the limited documentation around Bedrock AgentCore and the missing support by cloudformation. ## Accomplishments that we're proud of I am very proud of the resulting solution and the possibilities it brings at a very low cost and high scalability. ## What we learned I have gained a lot of experience and knowledge regarding new SDKs and frameworks. ## What's next for InsightModAI I intend to further develop the solution and bring more features to it as I believe it has the potential to bring real value to small to medium sized organizations.
Built With
- amazon-web-services
- bedrock
- cloudformation
- github
- javascript
- python
Log in or sign up for Devpost to join the conversation.