CortexClick today launched a content generation platform based on large language models (LLMs) that have been specifically trained to create documentation, tutorials and technical blog posts, including screenshots, for organizations building software.
Company CEO Evan Boyle said the overall goal is to make it simpler for organizations to create documentation using AI agents trained using multiple models that are capable of understanding the functions of the underlying code that has been created.
While there is no shortage of AI writing tools, the software-as-a-service (SaaS) platform created by CortexClick goes a step further by providing access to multiple agents that challenge the output created, said Boyle. In effect, one agent acts as an editor that challenges the veracity of the content created by another agent, he noted.
Additionally, CortexClick has provided integration with the widely used GitHub software repository in addition to making available a developer application programming interface (API) and software development kit (SDK) to make it simpler to integrate other platforms.
The company in addition to providing access to governance tools makes it possible to require the output generated to adhere to specific guidelines, such as the brand voice of the organization or other rules that organizations define, said Boyle.
Few developers and the DevOps teams that support them enjoy creating documentation. One of the issues that often makes technical support challenging is the lack of available documentation showing how an application functions. In the absence of that documentation, IT teams are forced to create support tickets that inevitably developers wind up needing to respond. Every minute spent responding to those requests is, of course, that much less time that could have been spent on writing code.
The CortexClick platform jumpstarts the documentation process by rather than requiring developers to start with a blank page instead generate content they can edit as they need, said Boyle. That’s critical because many developers are not naturally gifted writers who especially enjoy creating expansive amounts of content, he added. In many cases, the current level of documentation provided by software development teams today is so sparse as to be nearly useless.
At the same time, marketing teams that may not have much technical expertise can use the platform to also create technical content that development teams by more easily review, noted Boyle. That content in a world where most sales of software is driven by online engagement is critical, he added.
It’s not clear how many application development teams are already using AI tools to generate documentation. The challenge is most of those general-purpose platforms have limited to no understanding of technical jargon.
Of course, being able to generate more documentation doesn’t automatically make it useful. Developers will still need to carefully review documentation before sharing it with organizations that are consuming their software, However, with any luck the number of technical support calls that DevOps teams wind up fielding should decline as documentation becomes more accessible.
In the meantime, it’s become clear documentation has emerged as one of the best use cases for generative AI. The challenge is now getting developers to focus more on providing better documentation using AI tools that should reduce the level of drudgery far too many associated with performing that task today.
