India-based Postman has made a name for itself as one of the most popular platforms for building and consuming APIs since its founding in 2014, with 500,000 organizations currently using the service. However, like many other SaaS services, the company’s valuation today is reportedly down from its 2021 high of $5.6 billion. However, the company is now launching an AI agent builder that combines large-scale language models, and is trying to be one of the first to enter the AI field. The Postman API platform and a visual editor that integrates the two to help non-developers build and test AI agents.
At first, jumping into AI agents may seem like an odd choice for a company like Postman, but Postman CEO and co-founder Abhinav Asthana spoke with me before launching the agent builder. As I pointed out when we talked, for these agents to be helpful, you need to: You can interact with various services. That method uses an API.
“We actually started having customers ask us about this, saying, ‘Okay, we’ve got an agent story coming up. What are you guys doing about it? And we’re trying to help them piece this together. There are two types of APIs, whether you’re using a homegrown system or a third-party one. is the LLM itself, such as the OpenAI API or Anthropic’s API, or your own API. The second bit is the tools that the agents use. So we said, “Postman is helping developers do all kinds of things with APIs, so this is emerging as a pattern. We thought, ‘If there is, we should do something here,”’ Asthana explained.
He also noted that a growing number of Postman users are non-developers who want to use the service to build simple API-centric applications. “So when AI technology started emerging, we had to test it over and over again. We were like, ‘Okay, can this work?’ And the combination of no-code, AI and agents seems like a good combination,” Asthana said.
Postman’s API network already features an API hub with contributions from over 18,000 companies, which also allows users to build AI agents that can interact with these APIs. To ensure these agents work as expected, Postman has added testing and evaluation tools to the service to support models such as OpenAI, Google’s Gemini, Anthropic’s Claude, Cohere, and Meta’s Llama. ) and test the prompts and input. This allows users to use Postman Flows, the company’s visual development environment, to build these agents and interact with the API.
Currently, Postman’s AI agents are primarily focused on enabling backend operations. Although developers can use this service to power their front-end services, Postman itself does not yet provide a user interface to directly interact with these agents. Over time, this may change. “We are all very bullish on agent software, but I think the way we interact with these systems will also evolve,” Asthana said.
Of course, Postman isn’t the only company in this space, but Asthana is differentiating its service from others, especially the big cloud providers, by offering a better developer experience in addition to its existing API hub. I think it can be made into “In my opinion, cloud providers have always prioritized cloud usage over developer experience, and I think they’re happy to partner with any company that actually provides a decent developer experience.” he said.
The company is still figuring out how to best charge for this new service. Asthana said pricing will evolve over time as the team considers how to best determine the value these agents bring.