AI Engineering ROI: Moving From Token Maxing to Outcomes

Vitaly Gordon · Cofounder and CEO · Faros

Oct 2026AI Engineering Leadership

01 — The conversation

GuestVitaly GordonCofounder and CEOFarosProfile →

AI coding tools can help engineers write code faster, but faster code does not automatically produce better business outcomes. Vitaly Gordon, cofounder and CEO of Faros, joins The Tech Trek to explain why engineering leaders need to connect AI usage to measurable results, rethink enterprise processes, and reconsider what they need from AI ready engineers.

Key ideas

01AI spend can grow faster than the value teams can measure.

02Faster code generation does not remove bottlenecks elsewhere in software delivery.

03Enterprise AI adoption requires changes to workflows, security, and operating models.

04AI ready engineers increasingly need both domain expertise and practical AI skills.

05Faster engineering does not necessarily mean fewer engineers.

Vitaly explains why moving from token maxing to outcome maxing gives engineering leaders a better way to think about AI adoption and ROI. The conversation covers how token spend can scale quickly even when customer value and product output are harder to measure.

For large organizations, the challenge goes beyond choosing tools. Existing review processes, security requirements, legacy systems, and hiring needs still shape what teams can deliver. Vitaly also shares why he is skeptical that AI will simply eliminate large numbers of engineering jobs.