About This Architecture

Event-driven AI agent workflow orchestrates daily stock sentiment analysis from Reddit using AWS Step Functions, Lambda, and Amazon Bedrock. EventBridge Scheduler triggers at 11:59 PM daily, invoking Step Functions to coordinate three Lambda functions: Reddit Scraper pulls posts via Reddit API, Sentiment Analyzer processes text through Bedrock's LLM engine, and Report Generator compiles insights. Raw posts land in S3, sentiment scores persist in DynamoDB with composite keys (record_id, date, market), and final reports deliver to subscribers via SES. This serverless architecture demonstrates best practices for AI agent orchestration, separating concerns across orchestration, processing, storage, delivery, and governance layers. Fork this diagram on Diagrams.so to customize the workflow, adjust DynamoDB schema for your market segments, or swap Bedrock models for different sentiment analysis approaches.

People also ask

How do I build a serverless AI agent workflow on AWS for daily stock sentiment analysis from Reddit?

Use EventBridge Scheduler to trigger Step Functions daily, orchestrating Lambda functions for Reddit scraping, Bedrock LLM sentiment analysis, and report generation. Store raw posts in S3, sentiment records in DynamoDB, and deliver via SES. This diagram shows the complete event-driven architecture with governance layers.