C3、製造業・エネルギー業界向け専門知識搭載のエージェント型コーディングプラットフォーム「C3 Code」を発表

3 Min ReadEnterprise AI application vendor C3 AI on Wednesday introduced C3 Code, a new product that combines agentic coding with the C3 Agentic AI platform. C3 Code closely follows other successful coding tools such as Claude Code from Anthropic and Codex from OpenAI but might appeal to businesses seeking a domain-specific coding platform. C3 Code enables business analysts, developers and data scientists to build enterprise AI applications in hours using natural language, according to C3 AI. The platform has prebuilt data models for industries such as manufacturing, energy, finance, defense, utilities and healthcare. It also includes prebuilt machine learning models for enterprise use cases such as demand forecasting and predictive maintenance. According to C3, the tool is not locked into any large language model, so enterprises can pick whichever LLMs they want. C3 Code might seem like just another AI coding assistant. Still, the vendor is targeting businesses that are not interested in generic coding AI agents but instead want domain-specific knowledge embedded in their platforms. While generic agents like Claude Code or Codex might be helpful for certain businesses, they can also leave more room for error because they are trained on generic knowledge and finetuned to fit specific applications instead of being trained on domain knowledgeRelated:OpenAI Updates Agents SDK, Aims at Building Secure AgentsExpertise NeededSo, businesses might be looking to work with a vendor such as C3 AI that has expertise in manufacturing, supply chain, and energy."If you're going to do agentic and you're going to do it way to avoid possibly hallucinations, you need to know not just what the data in your database means, but how you arrive at it," said Bradley Shimmin, an analyst at Futurum Group. "This is a situation where expertise matters more than speed of innovation and capability."Not only does domain expertise matter, but prebuilt models can also be helpful for businesses that lack the in-house skill to build these agentic applications from the ground up, Shimmin said."Their platform is tailored to supporting [companies] out of the gate with ready-to-go capabilities," he said. "It's a shortcut to proven value, as opposed to having to either do it yourself or try to fix what you already have, which may not be agentic-ready."Lock-in and Other ChallengesHowever, while businesses using C3 Code might not be tied to a specific model, they may still be locked into C3 AI's ecosystem or platform, which includes its methodology, full-stack architecture and systems, Shimmin said.Related:As AI Infosec Woes Heighten, IBM Intros Autonomous Security ServiceBut some enterprises might not care. For example, for large enterprises, lock-in might not matter because they have the resources to adapt depending on what is favorable in the market, but for small enterprises that need to be flexible, lock-in might not be ideal."So, how much expertise do you need? How open do you really need to be? Are the questions that any company must ask themselves before they partner with a vendor," he added.Another challenge for C3 Code is that it's unclear how the vendor plans to connect enterprise data to its prebuilt models."This stuff, particularly connectors and such to data sources, doesn't happen with pixie dust," said Mark Beccue, an analyst at Omdia, a division of Informa TechTarget. He added that C3 did not provide some enterprise requirements for how it plans to connect data.for example, such as providing instructions that the data must be searchable and accessible using an API. "They were vague," he said.About the AuthorNews Writer, AI BusinessEsther Shittu brings four years of expertise covering artificial intelligence technologies and industry trends. As co-host of the "Targeting AI" podcast, she talks to thought leaders and practitioners exploring critical AI developments. Previous to AI Business, she wrote for several publications including the New York Daily News, Bklyner and the Brooklyn Daily Eagle. When she's not diving deep into the world of AI, she spends her time on passion projects and raising her three daughters.