Mike Vizard summarizes Harness CEO Jyoti Bansal’s vision for reimagining CI/CD and DevSecOps workflows using AI. The article explores automation of build, security, and developer operations with emerging AI-driven tools.

Harness CEO Advocates AI-Driven Transformation of CI/CD Workflows

Author: Mike Vizard

Key Highlights

  • Jyoti Bansal, CEO of Harness, calls for a fundamental reimagining of CI/CD workflows to capitalize on AI advancements, as described in his keynote at Unscripted 2025.
  • Traditional CI/CD and DevSecOps processes are overwhelmed by the surge in code volume produced by AI coding tools, with current pipelines struggling to handle increases up to tenfold in some organizations.
  • Bansal claims that 70% of build and deployment time is lost to manual or homegrown tasks, rather than development.

Harness’ AI-Centric Platform Features

  • AI-driven Issue Remediation: Modern CI/CD platforms can now use AI to automatically address build issues and remediate vulnerabilities, eliminating repetitive tasks and reducing friction.
  • Human-in-the-Loop Automation: AI not only detects tasks needed for successful builds but can execute them autonomously, while keeping developers updated and involved for key approvals.
  • Integrating DevSecOps: The Harness AI platform, leveraging tools such as Traceable, can write code to fix detected security vulnerabilities, which developers can then review and insert into upcoming builds.
  • Leverage of Large Language Models (LLMs): By employing LLM reasoning, Harness claims to accelerate root cause analysis and automate the generation and testing of code changes.
  • Internal Developer Portal (IDP): Harness also enhances developer self-service through an AI-powered IDP based on the open source Backstage platform.
  • Cost Optimization: AI agents can enforce FinOps policies to optimize operational costs within the CI/CD ecosystem.

Industry Context & Adoption

  • According to Bansal, more than 100 Harness customers have already adopted these AI-driven workflows to automate and scale development operations.
  • The article notes industry debate about whether organizations will require entirely new platforms to enable deep AI automation or if AI agents can be incrementally layered onto current tools.

Conclusion

Bansal concludes that engineering teams will inevitably rely more on AI-driven automation to manage software delivery at scale. The increasing integration of AI is shifting developer focus from repetitive tasks to higher-level responsibilities within modern DevOps and security workflows.

References and further reading:

This post appeared first on “DevOps Blog”. Read the entire article here