Last week Abhijit Kane, co-founder of Postman gave me a peek into how he builds with AI.
Here are 5 takeways from the live lesson on Maven:
1️⃣ From Developer Tool to AI-Native Platform With AI agents and their visual programming tool Flows, Postman is removing barriers that previously required deep technical knowledge and lets product managers, IT teams, and non-developers to quickly build integrations and automations.
2️⃣ The Power of Internal AI Agents Postman built internal agents like Aira that transformed how PMs and other teams consume their own data. Their GitHub issue tracker was always there, but it took AI to make it truly usable. Now employees ask the agent even the smallest questions through Slack rather than manually searching.
3️⃣ Rethinking Tool Architecture for AI PostBot -> Agent Mode illustrates a change in building AI products. Instead of creating fixed workflows with AI sprinkled on top, they exposed all Postman capabilities as tools and let the AI orchestrate them. This MCP-inspired approach unlocks workflows users couldn’t express through traditional UI.
4️⃣ Visual Programming Meets API Integration Flows lets teams automate mundane but necessary tasks like feature flag cleanup and data summarization without writing code. Many high-effort, high-impact tasks remain undone not because they’re impossible, but because they’re not worth traditional development time.
5️⃣ The Future is AI-Native UX If agents can understand natural language intent and orchestrate complex operations, why are we still clicking through navigation bars and tabs? Postman is betting that AI-native interfaces will blend chat with redesigned visual elements, moving beyond autocomplete and chatbots to something new. The PMs and designers that figure this out first will define the next generation of dev tools.
You can watch the full recording at: https://lnkd.in/g6jTvW9S
Let me know your own takeways!