Postman AI Agent Builder: Automate API Workflows
Postman Enters the AI Arena with New Agent Builder
Founded in 2014, Postman, the India-based company, has established itself as a leading platform for API construction and utilization. Currently, 500,000 organizations leverage its services. Despite a reported valuation decrease from its 2021 peak of $5.6 billion, the company is now significantly investing in artificial intelligence.
Postman has launched an AI agent builder, integrating large language models with its existing API platform. A complementary visual editor is also included, designed to empower both developers and non-developers to create and test AI agents effectively.
The Rationale Behind the Move
The decision to venture into AI agents might appear unconventional for Postman, but CEO and co-founder Abhinav Asthana explained the underlying logic. He highlighted the necessity for these agents to interface with diverse services, a process fundamentally reliant on APIs.
Asthana shared that customer demand played a crucial role in this development. Users began inquiring about Postman’s plans regarding AI agents, prompting the company to analyze existing workflows. It became clear that APIs were central to agent functionality, encompassing both the LLMs themselves and the tools agents utilize.
Expanding User Base and No-Code Integration
Postman has observed a growing segment of non-developer users who utilize the platform for constructing simple, API-driven applications. This trend, coupled with the emergence of AI technology, led to the exploration of a synergistic combination: no-code development, AI, and agents.
The company’s API network, boasting contributions from over 18,000 companies through its API hub, now facilitates the creation of AI agents capable of interacting with these APIs.
Testing and Evaluation Tools
To guarantee the reliability of these agents, Postman is introducing testing and evaluation capabilities. These tools will allow users to assess prompts and inputs for supported models, including OpenAI, Google’s Gemini, Anthropic’s Claude, Cohere, and Meta’s Llama.
Users can then leverage Postman Flows, the company’s visual development environment, to construct these agents and orchestrate their interactions with APIs.
Current Focus and Future Possibilities
Currently, Postman’s AI agents are primarily geared towards enabling back-end operations. While front-end applications are possible, the platform does not yet offer a direct user interface for agent interaction.
Asthana anticipates that the methods of interacting with these systems will evolve over time, expressing optimism about the future of agent software.
Differentiation and Developer Experience
Postman recognizes the competitive landscape in the AI space. However, Asthana believes its existing API hub and commitment to a superior developer experience will set it apart, particularly when compared to large cloud providers.
He suggests that cloud providers often prioritize cloud consumption over developer needs, creating opportunities for partnerships with companies focused on enhancing the developer experience.
Pricing and Future Development
The company is still determining the optimal pricing model for the new service. Asthana indicated that pricing will be refined as the team gains a better understanding of the value these agents deliver.
TechCrunch offers an AI-focused newsletter; subscribe here to receive it weekly.
