PostHog announced it is training its own AI models to power more proactive, self-driving products, the company said in a blog post. This initiative aims to build AI features that not only provide answers but also act on them and improve over time, with the first product under this vision, PostHog Code, now in beta (posthog.com).
Over the past year, PostHog integrated AI-powered tools such as an AI installation wizard, PostHog AI, and the Model Context Protocol (MCP) into its platform. The new effort involves training AI models directly on data within PostHog to enable more autonomous and intelligent product capabilities. CEO James Hawkins described this as the next chapter for the company, focusing on delivering solutions that proactively assist users rather than requiring manual input (posthog.com).
This move reflects broader trends in software development where AI-driven automation and self-service tools are becoming central to product strategies. By training proprietary models on its own data, PostHog aims to differentiate its offerings and enhance user experience. The beta launch of PostHog Code signals the company’s commitment to advancing AI integration in analytics and product management, aligning with industry shifts toward more intelligent, adaptive software (posthog.com).
PostHog plans to continue developing these AI models over the next six months, expanding their capabilities and integrating them into additional products. The company’s focus will be on creating AI features that not only surface insights but also take action and learn from outcomes, marking a significant evolution in its product roadmap (posthog.com).