Building AI Native Teams Through Shared Context

Eswar Vandenapu · Co-Founder & CTO · NovaWorks AI

Nov 2026AI Engineering Product Leadership

01 — The conversation

GuestEswar VandenapuCo-Founder and CTONovaWorks AIProfile →

Eswar Vandenapu, Co-Founder and CTO of NovaWorks AI, explains why becoming AI native involves more than adopting new tools. He shares how his team is building shared context into its workflows, rethinking agent design, and changing what it expects from engineers as AI takes on more development work.

Key ideas

01Shared context helps teams make decisions from a more consistent foundation.

02Agent design should account for the requirements of different environments rather than multiplying agents.

03Product specifications can become part of collaborative engineering reviews.

04Engineers increasingly need systems thinking and business knowledge alongside technical skills.

05AI generated plans still require people to question assumptions and exercise judgment.

Eswar describes NovaWorks AI's progression from introducing Claude across the company to using AI throughout its operations. As adoption expanded, the team found that people were making decisions based on different information and assumptions. NovaWorks began developing a shared context model managed through a Git repository, allowing teams to review changes and work from more consistent information.

The conversation explores why context matters more than agent count, how AI is bringing engineering and product responsibilities closer together, and why technical hiring is shifting toward systems thinking and domain expertise. Eswar also explains why teams should challenge AI generated plans and ask better questions rather than simply issue instructions.