The PodKnowledge|← All Episodes
Based on Lenny's Podcast data
Lenny's Knowledge Sketch

Products as Organisms,
Death of Org Charts

Asha Sharma
Chief VP of Product, Microsoft's AI Platform
AUG 28 2025
The Thesis

The Org Chart Becomes
the Work Chart

AI AGENTS DEPLOYED
"The way that you scale to that is with agents. When all of that happens, the org chart starts to become the work chart. You just don't need as many layers."
  • Products as organisms: self-adapting, learning, responding systems, not shipped features
  • Agents are the new employees: task-specific, always available, infinitely scalable
  • The death of the org chart: fluid team structures form around outcomes, not functions
  • "Humans will always decide in organizations how AI is used and what we want to apply it to"
Framework

The Agent-First Organization

TASKAGENTOUTCOME
15K
customers who have produced agents (Azure)
  • Agent types: research agents, writing agents, analysis agents, code agents, customer agents
  • Human roles in an agent-first org: direction, judgment, relationships, creativity
  • The coordination layer: how do humans and agents collaborate in real-time?
  • The accountability layer: who is responsible for what an agent produces?
From Asha"15,000 customers who have produced agents" on Azure. "The number of agents is actually millions."
Products as Living Systems

The New Product Paradigm

  • Old product: Ship feature → measure → iterate (quarterly cycle)
  • New product: Ship agent → observe behavior → adapt in real-time
  • The organism model: Products that learn from every interaction, continuously
  • What this requires: Real-time feedback loops, adaptive agents, judgment at the edge
The product PM in an agent world

PMs become systems designers and judgment setters. They don't manage features, they manage behaviors.

The accountability gap

When an agent makes a mistake, who is responsible? This is the hardest unsolved problem in agent-first products.

Playbook

Build Agent-First

  • Identify your 5 most repetitive team workflows, those are your first 5 agent opportunities
  • Design for human-agent handoffs: when does the agent escalate to a human?
  • Build accountability systems for agent output: who reviews, who approves, who is liable?
  • Redesign your product spec: what does the product do when users aren't watching?
Asha's warning"Companies fail when they're doing AI for AI's sake. They have a ton of projects kicking off at the same time without a blueprint, and they aren't treating it like a real investment, so they don't have the measurement, observability, and evals set up."
Contrarian

Agent-First Myths

✗Agents replace human judgmentINSTEAD →✓ Agents amplify human judgment. Humans decide what the agent optimizes for.
✗Org charts just need updatingINSTEAD →✓ Org charts need replacing. Fluid, outcome-oriented structures are the new operating model.
✗Products ship, then learnINSTEAD →✓ Products ship AND learn. Continuous adaptation is the new product lifecycle.
✗This is too early for most companiesINSTEAD →✓ 15,000 customers on Azure have already produced agents. Millions are running today.
Based on Asha Sharma's episode on Lenny's Podcast. All ideas on this page are from the episode.Watch on YouTube
GO DEEPER IN THE EPISODE
04:18 From “product as artifact” to “product as organism”06:20 The rise of post-training and the future of AI product development09:10 Successful AI companies: patterns and pitfalls12:01 The evolution of full-stack builders14:15 “The loop, not the lane”: the new organizing principle16:24 The future of user interfaces: from GUI to code-native19:34 The rise of the agentic society22:58 The “work chart” vs. the “org chart”26:24 How Microsoft is using agents28:23 Planning and strategy in the AI landscape35:38 The importance of platform fundamentals39:31 Lessons from industry giants42:10 What’s driving Asha44:30 Reinforcement learning (RL) and optimization loops49:19 Lightning round and final thoughts
Every sketch here began as a transcript.

Send us yours and we’ll draw it. $1 a transcript.

Sketch mine →
𝕏︎ X / Twitterin LinkedIn📸 Instagram🔗 Copy link
0:00