Engineer builds ai post-mortem bot with addams family creep factor to make failure reports deliciously macabre on aws lambda

Dario De Sio, Head of Software Engineering at Eleva and an AWS Certified Solution Architect Professional, has developed Nevermore.dev, a post-mortem generator utilizing AWS Lambda Durable Functions and Amazon Bedrock's Nova models. The tool aims to simplify the process of writing post-mortems by leveraging AI to generate clearer and more useful incident descriptions and root cause analyses. The system employs a large language model as a judge, evaluating outputs from multiple models to select the best result. De Sio's solution uses a single Durable Lambda Function to manage the workflow, eliminating the need for AWS Step Functions and reducing complexity. The function can handle parallel execution, checkpointing, and resumability, making it suitable for AI workflows. Nevermore.dev is deployed using AWS Amplify Gen 2 and CDK, with Amazon Cognito handling authentication and DynamoDB providing storage. The tool's development demonstrates the potential of Lambda Durable Functions for serious AI workflows, offering a more efficient and cost-effective alternative to traditional Step Functions.