LogLift

S3-Athena Log Analyzer - “no DDL, no SQL, just ask.”

@ AWS

GenAIAWSReactFastAPIBedrockAthena

Overview

AWS Support engineers spend significant time writing DDL and SQL just to inspect S3 logs. LogLift automates the entire pipeline - reducing average analysis time by over 30%. Started as a Streamlit prototype, then redesigned into a production-grade architecture with a typed API and React frontend.

Demo

Gallery

LogLift web UI — Natural-language log analysis UI

LogLift architecture diagram — RAG-based architecture - Bedrock, Glue, Athena

Highlights

Automatic table generation

Detects S3 log structures (S3 Access, CloudFront, ALB, VPC Flow, CloudTrail) and auto-creates partitioned Athena tables with the right schema.

Natural language → SQL

Engineers ask questions in plain language; Amazon Bedrock (Claude) generates optimized SQL with automatic error correction and retry.

AI-driven analysis & visualization

Anomaly detection, root-cause analysis, and automated dashboard generation on top of query results.

Stack

ReactTypeScriptFastAPIPydanticboto3Amazon BedrockAthenaGlueS3