July 2025 Summaries

Harness Database DevOps enhances the Liquibase developer experience by automating the changelog generation process, streamlining database change management, and integrating schema changes into a CI/CD framework. By treating schema changes as version-controlled code, it allows for faster and more reliable deployments, ensuring traceability, compliance, and governance. Developers can use Harness to automatically generate database changes through snapshots and diffs, eliminating the need to manually write SQL or YAML. The pipeline ensures consistency across environments with a GitOps workflow, and policies can be enforced to maintain compliance, such as index naming conventions. This automation reduces human error and accelerates the development process, allowing developers to focus on schema improvements while ensuring changes are audit-ready and policy-compliant. Stephen Atwell, who leads Harness's Database DevOps product, emphasizes the importance of such tools for improving the developer experience and modernizing database CI/CD processes.

Harness has received the inaugural WINvaluable Award from Wiz, highlighting their collaborative effort to enhance cloud security through the Wiz Integrations (WIN) partner ecosystem, which now boasts over 200 integrations. This partnership addresses the challenges of security in a fast-paced, cloud-native environment by integrating Wiz's high-fidelity security insights into Harness's CI/CD platform, offering developers prioritized and actionable alerts to mitigate the "wall of noise" created by early security embedding. With the Harness Security Testing Orchestration (STO), teams can leverage Wiz's scanning capabilities within their pipelines, enabling the identification of "toxic combinations" and enhancing pre-production security measures. Harness AI further augments this integration by providing developers with specific remediation guidance, streamlining the process from detection to resolution. This collaboration not only validates their commitment to delivering developer-centric security solutions but also empowers organizations to innovate swiftly and securely, ensuring a robust foundation for cloud security. Eric Minick, recognized for his expertise in software delivery, plays a pivotal role in advancing Harness's market solutions.

Harness CI and Code have introduced new features designed to enhance developer productivity by reducing operational noise and improving control, such as smarter builds, Kotlin test optimization, repo favorites, and tag protection rules. These updates aim to increase developer velocity while maintaining control and transparency, especially critical when managing numerous repositories. Key features include Tag Rules for Git tags, allowing fine-grained control and user group management, and Favorites for prioritizing important repositories. Security is enhanced through audit logs for force pushes to main branches, providing transparency and traceability. Build Intelligence now supports Maven, optimizing build times by skipping redundant steps, and Test Intelligence for Kotlin projects reduces CI times by running only tests impacted by recent code changes. These updates are part of a broader effort by Dewan Ahmed, a Principal Developer Advocate at Harness, to enable efficient, reliable software delivery. Ahmed, with a background at IBM, Red Hat, and Aiven, is passionate about open-source and advocates for underrepresented groups in tech, offering pro bono career coaching as part of his commitment to the community.

Harness Database DevOps offers a solution to the challenges of database changes in modern CI/CD pipelines by integrating both state-based and script-based migration strategies. This hybrid approach enables teams to define schema changes declaratively while ensuring controlled, versioned execution of scripts. The process leverages GitOps workflows and integrates with Liquibase OSS to automate schema management, providing a balance between simplicity in definition and safety in execution. Harness Database DevOps streamlines the deployment of database changes, enhances developer confidence, and maintains data integrity by automating script generation, testing, and rollback, thus resolving operational challenges and accelerating CI/CD pipelines.

Harness AI achieved a notable milestone by securing the #4 position on the SWE-Bench Verified leaderboard, demonstrating its ability to autonomously solve real-world software issues from GitHub. This achievement highlights the tool's capacity to address the bottlenecks in software delivery by integrating, validating, and deploying code efficiently. SWE-Bench Verified evaluates AI agents on their coding skills with 500 real GitHub issues across Python repositories, requiring them to understand, fix, and validate code autonomously in one attempt. Harness AI employs a modular architecture and Claude 4 Sonnet's "Thinking Mode" to enhance problem-solving through deep reasoning and adaptive planning, reducing errors like hallucinations. It utilizes a multi-agent system with a Build & Test Agent and a Fixing Agent to explore repositories, execute build commands, and dynamically validate fixes, proving its capability to handle complex software delivery beyond simple coding tasks. This success is seen as a step towards developing AI that excels in real-world engineering environments, with advanced tools and scalable architecture designed for modern software delivery challenges.

The complexity of many DevOps toolchains, resulting from years of ad-hoc additions, is causing significant challenges for development teams by creating brittle pipelines and slowing business velocity. Instead of introducing more tools, a strategic approach focusing on building "golden paths," embracing smart automation, and unifying the developer experience is recommended to streamline processes and reduce developer frustration. This complexity leads to manual rollbacks, security delays, and productivity issues, as highlighted by the 2024 State of Developer Experience Report. The root of the problem lies not in individual tools but in their overall architecture and misaligned incentives between platform and development teams. To address these issues, the article suggests implementing reusable templates, automating security as a feature, and unifying essential integration points. Harness, a platform designed around these principles, offers solutions like reusable pipeline templates, AI-driven deployment confidence, and built-in governance to simplify software delivery.

Chaos engineering is essential for building resilient systems, but scaling its practices often requires specialized expertise that can create bottlenecks within organizations. To address this, Harness has introduced the AI Reliability Agent, an experimental feature designed to automate chaos engineering processes using artificial intelligence. This agent provides experiment recommendations, remediation guidance, and is accessible to teams at any skill level, specifically working with Kubernetes infrastructures. By automating decision-making processes, the AI Reliability Agent helps overcome the challenges associated with expanding chaos engineering practices, making it easier for teams to perform effective experiments and implement fixes. It integrates seamlessly into existing Harness workflows and provides intelligent guidance based on environment monitoring data to recommend new chaos experiments with pre-tuned parameters. Although automation capabilities are powerful, it is crucial to validate AI-generated fix recommendations with human experts to ensure they align with specific systems and business requirements. The integration of AI into chaos engineering workflows without requiring new tools or processes makes it easier for teams to adopt resilience testing, offering a practical path for organizations to scale their resilience practices effectively.

Harness AI enhances DevOps automation by integrating smarter, context-aware workflows directly into the Harness Platform, enabling rapid software delivery with minimal complexity. The platform's new capabilities include natural language pipeline generation, AI-driven troubleshooting, and policy enforcement, all designed to adapt to specific organizational environments and standards. These features leverage large language models and a proprietary knowledge graph to help teams manage challenges such as fragmented toolchains and compliance requirements. With enterprise-grade guardrails and full auditability, Harness AI offers a streamlined approach that reduces manual tasks, accelerates onboarding, and improves issue resolution times. The seamless integration into the Harness Platform supports a wide range of DevOps activities, ensuring efficient and scalable software delivery.

Harness University's Instructor-Led Training (ILT) program is designed to provide hands-on, immersive learning experiences for technical practitioners, focusing on enhancing software delivery skills by leveraging the capabilities of the AI-enabled Harness platform. Having trained over 1,000 students across various cohorts, the program emphasizes the importance of continuous learning amidst the evolving landscape of software delivery. It offers both in-person and virtual training options, allowing flexibility for diverse teams, including global systems integrators and DevOps practitioners, to engage with the material in a format that suits their needs. The ILT program, led by Harness engineers, encourages interactive dialogue and peer learning, with a focus on practical application of Harness Modules such as CI, CD, and GitOps. Furthermore, the program complements Harness's free Self-Paced Training and Developer Certifications, offering a comprehensive path for individuals and organizations to enhance their software delivery capabilities.

Harness Infrastructure as Code Management (IaCM) marks its first year with significant enhancements focused on reusability and scalability, introducing features like Module Registry and Workspace Templates. These tools enable teams to standardize and reuse infrastructure effectively, accelerating delivery, reducing risks, and supporting scalable infrastructure automation without sacrificing governance or developer velocity. The Module Registry allows for centralized management of "golden templates," promoting consistency and compliance, while Workspace Templates offer standardized blueprints for new projects, streamlining onboarding and reducing setup errors. By embedding software engineering best practices such as the DRY principle into infrastructure management, Harness IaCM addresses common challenges like complexity and fragmentation and integrates with tools like Harness CI/CD and Cloud Cost Management for complete lifecycle automation. The solution has gained traction with enterprises, demonstrating strong capability in managing large-scale infrastructure through automation, reusability, and visibility, and plans are underway to further expand its support and standardization features.

Harness Templates offer a flexible solution for managing software pipelines by using the "Insert Block" feature, which allows platform engineers to maintain standardization while granting development teams the ability to customize specific pipeline steps. This approach addresses the common issue where rigid templates force teams to create separate, unmanageable pipelines when their needs diverge slightly from the standard. By incorporating controlled flexibility, such as allowing teams to choose their preferred security scanning tools within a standardized template, organizations can prevent pipeline sprawl and maintain governance. This method ensures that teams remain connected to the core template, benefiting from ongoing updates and improvements, while still meeting specific requirements. The "Insert Block" serves as a strategic compromise, enabling a balance between standardization and developer autonomy to enhance the adoption of robust and practical tools across the organization.

Harness's Cloud AutoStopping technology offers a novel approach to cloud cost optimization by employing intelligent, real-time automation to eliminate idle resources without disrupting workflows, addressing a significant challenge in cloud operations. This patented technology dynamically adjusts cloud environments based on actual demand, enabling organizations to significantly reduce waste—up to 70%—by making context-aware decisions about resource allocation, as opposed to relying on inflexible, time-based schedules. Particularly effective in managing non-production environments, Cloud AutoStopping ensures resources are only used when needed, thereby reducing unnecessary expenditure. This system is integrated with Harness's Cloud Cost Management platform, providing a comprehensive solution that maintains operational reliability without manual intervention. By aligning resource usage with actual demand across different time zones, it enhances productivity for globally distributed teams. As cloud-native architectures become more prevalent, Harness's innovation represents a shift towards more sophisticated, automated cost management tools, crucial for sustainable growth in modern software delivery.

Harness's new Chaos MCP Tools leverage AI Agents and AI Tools to facilitate resilience testing through natural language prompts, enabling users to assess the resilience of business-critical applications or services without needing extensive technical knowledge. These tools simplify the orchestration of chaos experiments by translating user prompts into API calls, making resilience testing more accessible amid the widespread digital transformation and cloud-native system adoption. The tools include features for listing, describing, running, and obtaining results from chaos experiments, helping users gain actionable insights into service resilience. The rise of AI and LLMs has reduced the skill barrier for conducting effective resilience testing, aligning with enterprises' growing need for robust resilience management. Harness has made its MCP server open-source, providing documentation and video tutorials to assist users in setting up and utilizing the server for resilience testing.

Harness Security Testing Orchestration (STO) integrates with Checkmarx One to enhance application security testing within CI/CD pipelines by automating scans for vulnerabilities and providing a centralized view of results. This integration allows DevOps and AppSec teams to embed security checks in their workflows, automatically evaluating code commits or builds for security issues using Checkmarx One's platform. The system normalizes and deduplicates scan results, enforcing governance policies such as blocking pipelines if critical vulnerabilities are detected. Benefits include early detection of security flaws, automated remediation suggestions, and policy-driven governance, ensuring comprehensive application security. This streamlined integration reduces overhead for development and security teams while maintaining robust security governance throughout the software delivery lifecycle.

Database deployments often struggle with common Git strategies, leading to issues like merge conflicts and broken rollbacks, which can hinder release processes. To address these challenges, the text advocates for a trunk-based, context-driven GitOps model, enhancing scalability, traceability, and release velocity. Traditional per-environment branching models, such as GitFlow, are criticized for their inefficiencies, while trunk-based development is lauded for its simplicity and reduction of integration friction. This approach utilizes a single mainline branch and environment metadata to manage database changes, promoting consistency and reducing errors. With tools like Harness Database DevOps and Liquibase, the approach emphasizes automation, observability, and seamless rollbacks, transforming database changes into reliable, repeatable processes akin to application code. The text highlights the importance of a clean Git strategy, not just for organization, but for ensuring resilience, safety, and speed in database deployments.