Seungeun Song

← Back to Projects

Sep 2025 – Dec 2025 (Irvine)

TrendPulse – AI Trend Intelligence for Travel Creators

An AI-powered trend analysis platform for travel content creators that collects online travel data, analyzes trends with AI, and delivers insights and content ideas.

Overview

TrendPulse is an AI-powered trend analysis platform designed for travel content creators. The platform collects travel-related data from online sources, analyzes trends with AI, and provides insights and content ideas to help creators produce timely content.

Tech Stack

AWS Lambda EventBridge Step Functions S3 Amazon Bedrock OpenSearch API Gateway Cognito CloudFront DynamoDB

Architecture Diagram

TrendPulse data pipeline architecture
Data Pipeline Architecture Scheduled collection, AI analysis, and OpenSearch indexing
TrendPulse user-facing architecture
User-Facing Architecture CloudFront web app with API Gateway, Cognito, and Lambda

Workflow

  • Scheduled data collection from online platforms.
  • Data storage and preprocessing in AWS.
  • AI-powered trend analysis and indexing.
  • Trend search and visualization through the web application.

Key Features

  • Scheduled trend data collection
  • AI-powered trend analysis
  • AI-generated content ideas
  • Travel trend search
  • AWS serverless architecture

My Contributions

  • Contributed to the design of the overall AWS architecture for both the data pipeline and user-facing services.
  • Designed a serverless data pipeline using EventBridge, Step Functions, Lambda, S3, Bedrock, and OpenSearch.
  • Selected AWS services based on scalability, maintainability, and cost considerations.
  • Estimated AWS infrastructure costs and evaluated deployment strategies for the project.
  • Participated in technical discussions on cloud architecture and system design.

Challenges

  • Designing an event-driven architecture capable of handling scheduled data processing workflows.
  • Selecting appropriate AWS services while balancing system requirements and infrastructure costs.
  • Structuring a modular architecture that separates data processing from user-facing services.

What I Learned

  • Learned how to design end-to-end serverless architectures on AWS.
  • Gained practical experience selecting AWS services based on scalability, maintainability, and cost.
  • Improved my understanding of cloud architecture design and event-driven systems.
  • Learned how to estimate cloud infrastructure costs during the system design process.

Demo Video

Frontend Demo
AWS Architecture Demo

Presentation

TrendPulse presentation slide 1
1 / 22