Continuous software delivery startup Harness Inc. said today it’s now able to harness the potential of generative artificial intelligence to help users automate many key aspects of software development.
The company says its AI Development Assistant will be integrated across all of its workflows and capabilities, including continuous integration, continuous deployment, cloud cost management and feature flags.
Harness is the creator of a continuous delivery-as-a-service platform that leverages machine learning algorithms to monitor new software releases in case they create problems for developers. The service makes it easier for developers to understand what’s happening in an application’s baseline environment, and if a new update starts acting up, they can initiate an automatic rollback to the previous version of that app.
The company said its AI Development Assistant is launching today with three specific features designed to alleviate key pain points in the software development lifecycle. For instance, it can assist with resolution of builds and deployment failures, so if an update goes wrong, it can analyze the log files and correlate error messages with known issues.
In this way, developers can use generative AI to troubleshoot any deployment failures much more quickly, Harness said, without needing to manually sift through millions of log lines. The AI Development Assistant can also suggest fixes and predict potential problems before the build is initiated.
Another feature pertains to automated vulnerability fixing, with the AI Development Assistant able to identify security vulnerabilities automatically and generate fixes. This feature integrates with the Harness Security Testing Orchestration module and is trained on all publicly known common vulnerabilities and exposures and common weakness enumerations.
Finally, Harness said its AI Development Assistant can simplify the process of creating and maintaining policies to manage cloud assets. Users can define policies for governing cloud assets and costs with natural language, automating what has traditionally always been a complex, manual process.
Harness said it expects its new service to boost the productivity of software engineering teams by around 30% to 50%, expanding the influence of AI beyond coding to every other facet of software development.
Constellation Research Inc. analyst Holger Mueller told SiliconANGLE that is a reasonable claim. He said enterprises are keen to activate their developer teams to build new generative AI applications, and so it was only a matter of time before the first frameworks emerged to make them more productive, using the very same technology. “Not surprisingly, this innovation comes from a startup, and it will be crucial to watch the initial uptake and see what kind of productivity boost it really provides to developers,” he said.
Harness added that it has also taken steps to ease concerns around data privacy, security and copyright infringement. The company explained that the AI Development Assistant’s stack, including its custom large language model, training data and vector database, are designed in such a way that all of the source code used in training has permissive licenses. It said this means its AI can be used without worrying that proprietary or copyleft code is used for training the Harness AI stack.
Jyoti Bansal (pictured), co-founder and chief executive of Harness, said his company’s goal is to break down barriers in software delivery, using AI to enhance the software development lifecycle for developers. “We’re committed to harnessing the potential of generative AI to address real-world developer challenges across the entire lifecycle of code,” he said.