More people came because the AI question belonged to more than one role.
About twenty attendees were expected for Howard Pearlmutter's custom AI-for-product-development seminar at Telos Alliance. Forty-seven came.
The audience included software and hardware developers, product and project leaders, executives, and skeptics. For an organization building combined hardware and software products for broadcast infrastructure, that range made sense.
The value of AI-assisted development could not be understood only through the act of generating code. It reached into requirements, prototypes, integration, learning, and the people expected to carry the result.
Begin with the organization's own work.
Howard prepared with leadership, including the CTO, VP of Engineering, and division leadership. The engagement addressed foundational knowledge, mental models, and deep insight into using AI coding tools effectively.
The opportunities included faster prototyping, coding, and testing, improved test coverage, and interface work. The discussion also had to make room for skepticism and the practical questions raised by adoption.
John Schur recognized Howard's attention to skeptical concerns and a careful path forward. Experienced people could bring their knowledge into the opportunity.
Practical prototypes supplied another kind of learning.
Howard built AI-assisted prototypes and demonstrations during the engagement. The work included Apple Silicon CPU/GPU execution, MPSGraph, Java native integration, Netty, Vert.x, Quarkus, and Jakarta EE CDI.
Dynamic Media Facility and Media eXchange Layer work connected enterprise software, high-performance computing, media, graphics, and simulation. It made emerging requirements inspectable at several levels of detail.
These were concrete examples of moving among layers: from low-level implementation to middleware to a domain-specific workflow. That breadth helped connect the AI discussion with actual engineering.
The client saw a combination of contributions.
Schur wrote:
“I felt that Telos received multiple times the value of our investment.”
His recommendation recognized requirements discovery, thought partnership, practical treatment of risk, and technical delivery. He also described Howard's mastery of AI tooling as evident in application, across implementation, languages, native integration, middleware, and media workflows.
For a team considering outside help, the combination matters. Learning and practical work can reinforce one another when the same contributor understands several sides of the problem.
What makes the example relevant to your team?
You may need more shared understanding before adopting a tool. You may need experienced skeptics involved earlier. You may need a prototype that helps several roles see the same technical question. Or you may need mentoring that connects a new capability with existing knowledge.
These are different purposes. Each can be served at a useful bounded scale.
Explore team workshops, adoption review, mentoring, or informative prototypes. Begin with what your people need to become more able to understand and do together.