The ultimate vision for AI in product isn't just generating specs. It's creating a dynamic knowledge base where shipping a product feeds new data back into the system, continuously updating the company's strategic context and improving all future decisions.

Instead of treating consulting and product as separate, CNX uses feedback from services projects to inform new features. A requested customization is often built directly into the core Valence product, benefiting all customers and creating a tight feedback loop.

To successfully automate complex workflows with AI, product teams must go beyond traditional discovery. A "forward-deployed PM" works on-site with customers, directly observing workflows and tweaking AI parameters like context windows and embeddings in real-time to achieve flawless automation.

Enterprises struggle to get value from AI due to a lack of iterative, data-science expertise. The winning model for AI companies isn't just selling APIs, but embedding "forward deployment" teams of engineers and scientists to co-create solutions, closing the gap between prototype and production value.

With AI agents automating raw code generation, an engineer's role is evolving beyond pure implementation. To stay valuable, engineers must now cultivate a deep understanding of business context and product taste to know *what* to build and *why*, not just *how*.

In AI, low prototyping costs and customer uncertainty make the traditional research-first PM model obsolete. The new approach is to build a prototype quickly, show it to customers to discover possibilities, and then iterate based on their reactions, effectively building the solution before the problem is fully defined.

Harvey's initial product was a tool for individual lawyers. The company found greater value by shifting focus to the productivity of entire legal teams and firms, tackling enterprise-level challenges like workflow orchestration, governance, and secure collaboration, which go far beyond simple model intelligence.

Avoid the trap of building features for a single customer, which grinds products to a halt. When a high-stakes customer makes a specific request, the goal is to reframe and build it in a way that benefits the entire customer base, turning a one-off demand into a strategic win-win.

Even at SpaceX, many engineers first heard from customers during a company all-hands. This feedback revealed the setup process was a huge pain point, leading to a dedicated team creating first-party mounting options. This shows that fundamental user research is critical even for highly technical, 'hard tech' products.