Dive Brief:

• Salesforce and SAP launched generative AI solutions for developers this week, making them the latest providers to point the nascent technology at the software creation process.
• Salesforce made its Code Builder product generally available Friday, enabling developers to code faster using generative AI to generate, test, debug and refactor code, the company said.
• SAP released Build Code solutions to streamline the process for developers using the company’s low-code tools, the company announced at its TechEd event Thursday. The capabilities are partially powered by the company’s generative AI copilot, Joule, assisting developers with data models, application logic and test script creation. 

Dive Insight:

Generative AI updates have come in waves. The first cycle tackled generalized capabilities as an intro to the technology, but vendors have turned their focus to tailoring solutions for specific job functions and industries. 

Other technologies have followed similar patterns. Cloud also started as a technology for general use, later customized by providers to meet the needs of specific industries. As early adopters and experimenters sought out valuable use cases in generative AI, aiding developers and the development lifecycle seemed promising. 

Making coding faster was already a target for popular tools, like GitHub’s Copilot, but the wave of generative AI interest reinvigorated the interest in coding assistance. Despite the potential opportunity, enterprise technology leaders haven’t yet gone all in on integrating generative AI into developers' day-to-day, but there is an appetite. 

More than half of developers said their organizations are interested in AI-powered code generation and code suggestions, according to a September GitLab report.

To meet potential demand, a growing list of providers have added features and enhanced generative AI coding capabilities. 

Anthropic and OpenAI expanded their models’ coding capabilities. Microsoft deployed GitHub Copilot for Business on Azure. Google Cloud and AWS also have tools and features related to coding assistance. 

Meta positioned its large language model, Code Llama, as a specialized coding tool free for research and commercial use in August. The LLM was trained on code-specific databases and supports popular languages. 

Google labeled its generative AI tool Bard as a programming assistant in April, though the company emphasized the need to double-check the tool’s responses for errors, bugs and vulnerabilities.