Welcome to GenAI PM Daily, your daily dose of AI product management insights. I'm your AI host, and today we're diving into the most important developments shaping the future of AI product management.
LangChainAI launched Enterprise Deep Research, a multi-agent system on LangGraph for automating deep enterprise research with real-time streaming and human-guided steering. The team also released Chatsky, a Python dialog framework with a dialog graph system built on LangGraph, and a production-ready solution for integrating private LLM APIs into LangChain and LangGraph applications with authentication, logging, and state management.
Shifting to AI tools, Aravind Srinivas added Perplexity Finance to his sidebar for one-click daily market insights. Meanwhile, Teresa Torres showcased ZenCity’s orchestration layer that reads agendas and auto-generates department-specific council meeting prep packets, saving hours of prep.
On the product side, Guillermo Rauch emphasized that shipping is a full-spectrum skill covering design, QA, marketing, and iteration, and said AI will push PMs to excel beyond coding. In related strategy insights, Aakash Gupta defined product sense as the ability to find the right solution for users and the business despite limited or ambiguous information. In related insights, George Nurijanian outlined an eight-point toolkit for platform PMs covering systems thinking, API design, DevOps fundamentals, data architecture, change management, and cross-team collaboration.
In industry news, Andrej Karpathy published a deep dive into a PyTorch MPS backend bug and speculated on when LLMs might fully automate debugging tasks. Separately, Logan Kilpatrick clarified the mental model separating Gemini as a personal assistant from AI Studio as a builder platform and hinted at integrating a Gemini Apps Canvas.
Moving into video insights, Wade Foster of Zapier mapped the AI automation spectrum—from deterministic workflows to agentic chatbots—showing that solutions in the middle balance determinism, cost, and LLM reasoning. He demonstrated a Zapier email-categorization agent using natural-language prompts to sort over 100 daily emails into action, executive, customer-insight, and informational buckets, trimming his inbox to under ten messages. It also highlighted that successful adoption relies on hackathons, demos, and trust-building through low-risk tasks.
Finally, Dhanji Prasanna, CTO of Block, explained how he transformed Block into an AI-native enterprise with Goose, an open-source agent built on the Model Context Protocol. Goose automates tasks across systems—from writing SQL in Snowflake to emailing PDF reports—saving teams eight to ten hours per week and cutting manual workloads by up to 25 percent. Prasanna also reorganized the company into a unified functional structure, aligning engineers and designers under single tech leaders to accelerate AI development across Square, Cash App, Afterpay, and Title.
That's a wrap on today's GenAI PM Daily. Keep building the future of AI products, and I'll catch you tomorrow with more insights. Until then, stay curious!