@TDep-SubstackPost (GPT5): Please read the text of the podcast transcript in the prompt and write a short post that summarizes the main points and incorporates any recent news articles that provide helpful context for the interview. Please make the post as concise as possible and avoid academic language or footnotes. put any linked articles or tweets inline in the text. Refer to podcast guests by their first names after the initial mention. Light editing and formatting for Substack.In this episode, George Robson and Pat Grady sit down with Jan Oberhauser, founder and CEO of n8n, to unpack how the company quadrupled revenue in eight months by shifting from workflow automation to an orchestration layer for AI applications. The throughline: go horizontal, bet on community, and make it easy to connect “everything to anything.”
From features to the value chain
Jan Oberhauser explains that the breakthrough came when n8n moved beyond bolting AI features onto workflows and instead enabled people to build full AI agents without needing Python. That reframed n8n as part of the AI value chain, not just an AI wrapper around an automation tool.
Context: The “orchestration” layer matters more as the stack fragments—foundation models, vector databases, tools, and protocols. Pinecone’s shift from “a vector DB” to “AI data infrastructure” (and its $100M Series B) was an early tell for Jan that infrastructure positioned in the AI value chain accrues outsized value.
What’s working
Community over leads: Addressing go-to-market, Jan says n8n deliberately dropped lead targets to focus on top-of-funnel adoption: empower the community, seed content, and let bottom-up usage compound. They optimized for large-organization adoption via community rather than gated funnels—an unusual but, in hindsight, decisive move.
Content drives adoption: The role of first party and community content, particularly YouTube how-to videos, has made n8n more discoverable to AI early adopters.
Fair-code, not “open source”: Jan is blunt about licensing: n8n’s source is open to use and self-host, but you can’t commercialize it. That consistency—never calling it OSI open source—builds trust and avoids the midstream license switches that burn communities. The ethos is open; the usage is sustainable.
The diversity of builders on n8n: Jan says usage quality held even as growth exploded. Builders are either technical or so motivated by a use case that they learn to code along the way. Most-used integrations still include Google Workspace and comms apps like Slack and Telegram—the natural front doors for AI agents—plus databases and internal tools.
Agents are getting easier—and more necessary
Jan argues the biggest unlock was letting builders compose real agents—prompt chaining, tools, vector retrieval, output parsers—without glue code. As agentic patterns spread, the need to coordinate models, tools, and data rises.
Context: Foundation models are still improving but the step-change pace has cooled; the action is shifting to reasoning, tool use, and agent systems. OpenAI’s o1 models emphasized reasoning and tool-use patterns. On the “open” side, Meta’s Llama 3/3.1 and Mistral’s open-weight releases show enterprises care as much about control and data locality as they do about license costs.
Why horizontal wins (again)
Pat asks whether vertical AI apps or horizontal platforms prevail. Jan’s view: verticals will bloom, but complexity forces orchestration back in—much like SaaS sprawl created the need for automation platforms. n8n’s advantage is neutrality: any model, any memory, any app.
Context: The market is standardizing on ways for models and tools to talk to each other. Anthropic’s Model Context Protocol (MCP) is becoming a common “HTTP for AI workflows,” enabling agent-to-agent interactions and plug-and-play tools. That strengthens the case for orchestration layers.
Product philosophy playbook
Balance free and enterprise users: Free usage drives product quality and long-term revenue; enterprise features serve a subset. n8n leans into free to capture the market, then meets enterprise needs as timing dictates.
Developers become “enablers”: In an agentic world, engineers set guardrails and empower end-users to build. That’s n8n’s sweet spot.
Short-term breakout area: AI-powered internal tools, where teams can take more risk and see faster ROI.
Jan has the ambition to be “Excel for AI” If everything goes right, he wants n8n to be the default answer when anyone thinks about building with AI—ideation to prototype to production—just as Excel became synonymous with spreadsheets.
Listen for the playbook: become indispensable infrastructure, nurture the community flywheel, and stay neutral so builders can pick the best model, memory, and tools for each job.
Hosted by George Robson and Pat Grady, Sequoia Capital
Mentioned in this episode:
Model Context Protocol (MCP): Open protocol that lets AI models safely use external tools and data that is used extensively by n8n for orchestration.
Vector database: A database optimized for storing and searching embeddings, now widely used for RAG systems in AI.
Granola: AI productivity tool mentioned by Jan as a recent favorite.
Her: A film that Jan says, “a few years ago, it was sci fi, and it’s now suddenly this thing that is just around the corner.”
