San Francisco-based artificial intelligence search company Perplexity is developing an internal coding assistant codenamed Teammate, signaling a strategic expansion into the competitive AI software development sector. Internal engineering teams have been utilizing the tool since May to manage long-horizon software projects, including debugging internal systems and monitoring services. According to internal communications, the platform is designed to oversee software lifecycles from inception to deployment rather than functioning as a simple autocomplete utility. The model-agnostic architecture allows Teammate to operate independently of any single large language model, a technical decision aimed at maximizing flexibility and reducing vendor dependency. Perplexity leadership has actively championed this transition, with Chief Technology Officer Denis Yarats recently instructing the engineering staff to gradually phase out manual code review in favor of AI-driven workflows by year-end. Yarats also addressed prevailing industry skepticism regarding AI-generated code quality, asserting that rigorous testing protocols will eliminate low-effort output typically criticized as slop. Though a public release timeline remains undefined, the initiative positions Perplexity, currently valued at $20 billion, to directly challenge established players in the AI coding arena. The move narrows the competitive gap with Cursor, Anthropic, and OpenAI, all of which have already deployed widely adopted developer tools. As the startup pivots from its core search engine operations toward comprehensive development infrastructure, industry observers anticipate that Teammate eventual market entry could significantly reshape the developer tooling landscape. Perplexity has declined to provide further commentary regarding the project commercialization strategy.