AI Briefing
KO

Hyundai AutoEver's Productivity-Boosting Hackathon Using GenAI Sandbox: A Success Story of Innovation and Collaboration

·2026.05.22 11:08

Key point

Hyundai AutoEver led AI proof-of-concepts by 14 teams through a GenAI Sandbox and a 4-week hackathon.

1 / 2

Details

Hyundai AutoEver built a GenAI Sandbox that lets employees experiment immediately, and based on it ran a roughly 4-week hackathon to drive company-wide GenAI adoption. A total of 14 teams, 150 people took part, validating work automation and operational efficiency gains.

The Sandbox was designed around 5 axes: security, cost, permissions, network, and development packages.

  • Security: The security framework was pre-configured with AWS Config, Amazon Macie, GuardDuty, Inspector, IAM Access Analyzer, and Security Hub.
  • Cost: Costs are automatically aggregated per project, and idle resources and excessive usage are detected and controlled.
  • Permission: Multi-tenant isolation within a single account was implemented using AWS IAM Identity Center and ABAC.
  • Network: Connection with the corporate IDC enables direct integration with in-house systems.
  • Development Package: API Gateway, Lambda, Bedrock Knowledge Base, OpenSearch Serverless, S3, and more are automatically configured so development can start immediately.

The hackathon produced real work-oriented projects, including automation of customer incident reports and Agent AI-based automation of big data platform operations. Participants reported reducing processes that used to take 2 to 3 hours of manual work down to 10 seconds, and the winning projects will move on to production support and further hackathons.

Adoption has since spread rapidly. So far, 25 offices and 30 teams have participated in proofs of concept. Automation of vehicle SW diagrams cut working time to under 20% while increasing output by about 3x. In operational incident analysis, automatic log analysis and improvement suggestions have significantly reduced the response burden.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.