This project is a cloud-native log analytics solution that ingests logs in real-time, stores them reliably, applies AI anomaly detection, and exposes them via an API and a React + TypeScript dashboard.
- Real-time log ingestion with Kinesis Firehose
- Scalable storage in S3 (raw logs) and DynamoDB (metadata & indexing)
- Amazon Bedrock Summarization to resume complex logs into simple text
- API Gateway with Lambda backend to expose logs and analytics
- React + Vite + Tailwind dashboard to visualize logs and anomalies
Data Flow:
- Applications / services generate logs.
- Logs are ingested via Kinesis Firehose.
- Firehose delivers logs to:
- Amazon S3 (long-term raw storage)
- DynamoDB (metadata for APIs)
- Amazon Bedrock detects anomalies and stores results in DynamoDB.
- API Gateway + Lambda expose endpoints to query logs.
- React Dashboard fetches logs from APIs for visualization.
.
├── terraform/ # Infrastructure code
├── lambda_functions/ # API layer (Lambda functions)
└── cloud-log-viz/ # Dashboard
- Install Terraform.
- Configure AWS credentials:
aws configure
- Deploy infrastructure:
cd terraform terraform init terraform plan terraform apply
This provisions:
- Kinesis Firehose
- S3 Bucket
- DynamoDB Table
- 2 Lambda functions
- API Gateway endpoints
- Go to dashboard:
cd cloud-log-viz - Install dependencies:
npm install
- Run the development server:
npm run dev
- Open http://localhost:5173.
Set in .env (for backend & dashboard):
AWS_API_GATEWAY_URL=https://<your-api-gateway-url>/prod- List all logs:
curl https://<api>/prod/logs
- Get a specific log:
curl https://<api>/prod/logs/{log_id}
