The advent of agentic AI is significantly reshaping the landscape of software development and product design. Agentic AI, with its autonomous characteristics, differs from typical AI, introducing challenges in governance and user trust. DataRobot, a platform specializing in AI, is adapting to these changes by offering agent and application templates. Instead of providing raw components, they now offer pre-built kits, fostering AI practitioners' work. This transition enables users to create production-quality applications rapidly through agent-generated interfaces. This shift demands a reevaluation of control distribution between users and agents, necessitating careful design for human-agent collaboration. The article draws parallels with AutoML, highlighting the need to balance automation with user agency over core tasks. DataRobot is expanding its user base to include agents, necessitating a shift in design methodologies to accommodate their needs. Traditional product design is evolving, where fixed solutions are replaced by dynamic, context-aware systems. Product teams' roles are shifting from building individual products to architecting systems. This change is challenging interaction design, as conventional affordances require new forms to adapt to on-demand interface generation. AI tools are addressing these issues by using established UX frameworks. The ultimate goal is designing systems where agents and humans work together effectively.