Founder StoriesShare Your Story
fin𝕏◎
GlobalFounderTimes
Stories of founders. Ideas that inspire.
Where Innovation
Meets Impact.
Submit Your Story
Have a founder story?

Share the problem, the decision and the result.

Submit Your Story
Founders · News analysis · GlobalFounderTimes Editorial

Abhinav Asthana’s Next Question for AI Agents: Who Is in Control?

Published October 25, 2026 · Coverage: August 2026

Event / announcement: August 18, 2026

Postman’s co-founder uses an August strategy statement to move the conversation from deploying agents to governing their actions.

Postman co-founders Abhijit Kane, Ankit Sobti and Abhinav Asthana. Archival official press-kit photograph.
Postman co-founders Abhijit Kane, Ankit Sobti and Abhinav Asthana. Archival official press-kit photograph. Credit: Postman press kit.

The most impressive agent demonstration is usually the one where software acts without being asked twice. The next question is less theatrical: who authorised the action, what could the agent access and how would someone stop it? That is the business problem Postman co-founder and CEO Abhinav Asthana addressed in an August strategy post.

Writing on August 18, Asthana introduced Postman.ai as a place for the company’s AI vision, research and benchmarking. He described an Agent Stack spanning API access, gateways, agent operations, documentation and the Postman platform. The named components included Passport, Fabric, Astro AI and Fern.

The post framed agent readiness around trust, access, governance, cost and operations. It is a founder’s account of the company’s direction, rather than independent evidence that the full approach solves those problems for every customer. It also does not remove the need to assess each product’s availability and limitations separately.

GFT analysis: this is a useful example of a founder extending a company’s established category into an adjacent problem. If agents take actions through APIs, the infrastructure that exposes and controls those APIs becomes strategically important. The argument connects an existing customer need to a new kind of user: software acting for people.

For smaller businesses adopting agents, the practical lesson is to define control before adding autonomy. Write down which action can happen automatically, which needs approval, who owns the credential and what record remains after execution. An agent that drafts a message and one that sends it create different obligations, even if the interface looks identical.

The commercial test for this direction is whether customers can operate useful agents with less uncertainty and administrative effort. More dashboards would not, by themselves, deliver that result. Asthana’s August statement identifies a real set of questions for the emerging market; the strength of Postman’s answer will depend on what customers can reliably do, observe and reverse in practice.

Sources and reporting

Independent editorial coverage based on public sources, with GlobalFounderTimes analysis. No interview, sponsorship, partnership or company endorsement is implied. Company claims are attributed to their source.

More from August 2026