DEV Community: Maxime Lenne

Nord Star Product 2026, Aurora edition: my takeaways from the 7 talks

Maxime Lenne — Fri, 30 Oct 2026 14:00:00 +0000

“We didn’t save time to go faster, but to spend more time on what really matters.” Anthony Lefebvre and Mathéo Evrard said it on the Nord Star Product stage on 22 October 2026, and it sums up the day quite well. Everyone talked about AI. Almost nobody said it would make decisions for us.

Here is my recap of the Aurora edition: the programme, the ideas and quotes from each talk, then what I take away from it, on the tech side and the human side. The talks were in French; the quotes below are my translations.

Nord Star Product, Aurora edition

Nord Star Product is the design and product day in Lille. It is organised by the Nord Star Product community and EuraCreative, which supports cultural and creative companies on the Plaine Images campus.

Date : Thursday 22 October 2026.
Venue : Plaine Images, Tourcoing, near Lille.
Format : 45 minutes of talk in the room, then a 15-minute break outside, because “the best moments of a conference happen at the coffee machine”.
In the evening : an after party with pizzas, craft beers, cocktails and a DJ set by Tremolo.

Time	Speakers	Talk
09:00	Arthur Hilhorst (Lovable)	Too fast to think
10:00	Noé Chagué and Maxime Podgorski (Éno)	The future of design with agentic AI
11:00	Anthony Lefebvre and Mathéo Evrard	We broke the Double Diamond!
13:30	Karine Dusart and Renaud Dorizon	A CPO’s first 60 days (panel)
14:40	Anicet Bart (Solfy)	Now everyone can build. But how do you win and keep users?
15:15	Stéphane Cruchon	The Design Sprint in the age of AI: deciding fast, and right
16:15	Roger Ingouacka (Petit Bambou)	Lessons learned: the life of a real Product Builder!

What I take away from each talk

Arthur Hilhorst (Lovable): too fast to think

Arthur is a product designer at Lovable, the Swedish start-up that builds applications with AI: about 60 million projects created, more than a million a week, and close to a billion visits a month on the generated apps. His talk came down to three words: speed, context, intent.

Speed. He asked the room to pick between thinking and speed: barely 20 to 30% of hands went up for thinking. Everyone feels the tension.
Context. Running 10, 20 or 30 agents in parallel creates a huge cognitive load. At Lovable, a “second brain”, an internal LLM connected to the codebase and to user feedback, holds the memory for them.
Intent. AI tools push us into the solution space too quickly, skipping the exploration of the problem.

On the organisation side: designers ship straight into the codebase, 70 to 90% of pull requests are generated by AI, and there is no design system in Figma any more. The source of truth is the code.

The part that struck me most was about cognitive load. We used to juggle Jira, Miro and Figma. Now we juggle 10 tabs, 10 agents and 10 contexts to keep in mind. Arthur has tried it: you can run 20 or 30 agents, but not do deep thinking over several days at the same time. Agents generate so much, so fast, that your brain can no longer keep track and drops out. And losing control is exactly when you start skipping the details. Hence the “second brain”: an external memory that holds the context for you. With one limit, given in answer to a question from the room: too much context gives generic answers, so start from one source and balance it afterwards.

“You only have one brain, so you can’t do everything.”

“If everyone starts at the same level, without intent, things really start to look alike.”

“If you wonder what craft and taste mean, look at Apple.”

His example stuck with me: the flyers in the supermarket, all generated with ChatGPT, where the protest poster looks like the piano lessons ad. Craft and taste will make the difference, hence a role emerging at Lovable: artistic profiles who understand code and iterate thousands of times on a visual detail.

Noé Chagué and Maxime Podgorski (Éno): an “AI ready” design system

Éno helps companies automate their business (invoicing, website and more). Its design team works directly in the code. Their case study: the 17 November launch week, with 143 screens generated in one week , about sixty of them in less than 24 hours.

The starting point: our first experiences with agents are disappointing. Components recreated, tokens ignored, spacing rules forgotten. For the Éno team, an agent asks itself three questions as soon as you give it an intent: what do I use? What am I allowed to do? And what do I do when the answer doesn’t exist? The third one is what matters.

Their answer:

Components are data : name, type, use case, description, state and status (usable by agents or not).
Semantic tokens , because that’s what the agent actually uses, and two MCP servers: one to find components, the other to find saved inspiration references.
Everything is a pull request , 100 to 150 a day. CI checks quality, design system compliance, regressions, metadata and the changelog entry. It’s the last line of defence.
When a component is missing , the agent opens the issue itself, assigns it, and handles it alone if it’s simple.

The outcome of the week: 80% of the screens shipped after review, 20% revealed gaps in the system. And those 20% were the most valuable part.

“What’s interesting isn’t the 140 screens shipped, it’s the gap, where the system fell short.”

“Once something is clear to an agent, it’s clear to a human too.”

“Good designers have years ahead of them.”

Their maturity scale is simple. Level 1: the agent guesses which component to use. Level 2: documented metadata, tokens and intents, the agent follows your conventions. Level 3, “AI ready”: the system talks to you and tells you when it falls short.

Anthony Lefebvre and Mathéo Evrard (Kiabi): we broke the Double Diamond

The topic came up by chance, in a corridor: Kiabi’s second-hand clothing collection, which works very well in stores and in the app (four times the web volume), had to evolve so customers could send their clothes from home. The usual answer came: “we don’t have time”. They set themselves the challenge of answering it in less than 48 hours.

The compressed Double Diamond: six phases, sixteen hours of work over 48 hours.

Understand and model (2 h): a workshop with the business, three possible journeys, only one kept, the simplest for users and for developers.
Prototype (7 h): Claude connected to Figma, with a year of mock-ups and a solid design system as its knowledge base. An in-house Figma plugin acts as UX writer. The “ready to dev” prototypes are deployed on Vercel.
Test (4 h): a store visit on 9 July, 5 customers and 3 employees.
Synthesise (2 h): 57 verbatims coded by AI (hypothesis, research question, signal) in a Notion database, then fed back into the Figma mock-ups.

80% for AI (screens, questionnaire, synthesis), 20% for the team: the details that make the difference.

The most interesting part is what AI did not compress. Three times, the clock slowed down: choosing the direction, deciding between options, going to see people. A quarter of the time, and most of what they learned.

“We’ve all used that excuse, but today, with AI, ‘we don’t have time’ doesn’t necessarily hold any more.”

“AI only compresses what is already solid.”

“We didn’t save time to go faster, but to spend more time on what really matters.”

Karine Dusart and Renaud Dorizon: a CPO’s first 60 days

A panel on a topic rarely covered at conferences: bringing a product culture into long-established mail-order companies, where “for a very long time, the product was the catalogue”. Renaud works at La Redoute, where product reports to marketing; Karine reports to an IT department.

The first 60 days. In theory: observe, listen, treat your team as a product. In practice: HR, budget and operational decisions to make from the first week. Three values set from day one: kindness, transparency, trust.
Project or product? No opposition. Big projects reassure shareholders and defend budgets; the challenge is to bring discovery and vision into them.
Marketing or IT? Marketing gives you the budget levers, IT gives you proximity with tech but makes it harder to engage the business. There is no universal right answer.
Hiring. For a PM: soft skills, posture and command of their insights rather than hard skills alone. For a designer: a product vision beyond the mock-up, and a tech sensitivity that has become essential.
AI. Security, licensing and budget constraints make it impossible to move as fast as the morning’s demos. At La Redoute, one PM is now a full-time AI builder, with ambassadors in every team.

“The business is going to be your BFF” (your best ally).

“Culture isn’t built in a month, maybe not even in a year. It’s what remains once you’ve left a company.”

“We don’t push mock-ups, we push experiences.”

“Accepting, as a leader, that my team sometimes loses time exploring AI. And I’m not afraid to say it.”

Karine also quoted Marty Cagan’s Transformed: the most important thing is not adopting a process, but earning the business’s trust. And Renaud shared his prioritisation filter: “what is my added value on this topic?”. His direct reports are there so he can trust them.

Anicet Bart (Solfy): everyone can build, now what?

Anicet founded Solfy, a music theory learning app incubated on the campus. His message: building is no longer the problem, finding and keeping users is.

Put the product in users’ hands from the first weeks.
Support them: a new product needs time to be adopted, and support is fundamental.
Show concrete uses, especially on Instagram.
Own the marketing: more than a third of the project budget goes to promotion.
Go and meet people, at trade shows, on the phone, all the time.
Structure the data the product generates from the start: it’s an asset AI can use tomorrow.

On the business model: a €40 yearly subscription per family, which local councils can pay for.

“You don’t retain a user. You make them succeed.”

Stéphane Cruchon: the Design Sprint in the age of AI

Stéphane has run Design Sprints since 2015, alongside Jake Knapp, who invented the method at Google Ventures. More than 100 companies coached in ten years, from start-ups to large groups.

His question: why still block five days when AI generates a prototype in a few minutes? Because a new project drowned in calendars dies, and because decisions are made by humans, not by agents.

Days 1 to 3, strategy : human friction, AI stays out. Everyone sketches their solution on paper. Those who use AI often come up with overloaded concepts and lose to three clear sticky notes.
Day 4, prototype : designers and agents lay the groundwork in the morning, the team makes the prototype real in the afternoon, with real data and the domain expert at the microphone.
Day 5, tests : real humans, non-negotiable.
After the sprint : AI transcribes and analyses, alongside the scorecards filled in by the team.

“It takes less than a minute to make something, but it takes at least five days to make a real business decision.”

“Days 1 to 3 are context engineering. Done by humans.”

“Design isn’t dead: you need it to decide, and to believe in the vision.”

His examples: an agent that analyses customer service calls and fills in the CRM, prototyped in a day and launched in less than six months; a heritage brand renamed two weeks after a sprint, because the tests showed the name was a problem; a Swiss neobank that grew to 800,000 customers. Each time, the emotional value of the process matters: people defend what they built together.

Roger Ingouacka (Petit Bambou): the life of a Product Builder

Roger, a developer turned Product Builder, has spent two years on Petit Bambou’s new vertical: Mouvement , released in September. The observation: people in their fifties in pain everywhere, with no condition, just after 15 to 20 years of sitting down. The goal: get your body moving every day, with no pressure to perform or to look a certain way.

From 2D to 3D. 2D drawing took too long to produce. The team switched to motion capture: an osteopath in a sensor suit for a full day, and local freelancers for the avatars.
A library of about 250 movements , cut second by second, with animated transitions between postures.
The vision : generate a session on demand. “My back hurts, generate a 3-minute session for me.”
The avatar choice over AI-generated video: a character whose face doesn’t change from one session to the next, and new movements other health professionals can record tomorrow.

All this with a team of about twenty people, without dedicated squads, and a very fair question from the audience: how do you stop an AI from chaining dangerous movements?

The threads running through the day

1. Humans keep what can’t be compressed. Kiabi measured it, Stéphane Cruchon built it into his method: choosing, deciding and going to see people don’t compress. AI takes the rest.

2. The source of truth becomes the code, and context becomes a product. Lovable and Éno no longer have a design system in Figma. Their design system talks to agents, and context is built like a tool: second brain, MCP, CI, changelog.

3. Speed without intent produces products that look alike. Craft, taste and decision-making become the difference again.

4. Building is no longer enough. Anicet Bart said it bluntly: now that everyone can build, the challenge is to win users and make them succeed.

How I felt about it

Very few tech people in the room. And yet, at a time when AI augments our skills and widens what we can do, these topics concern us more and more: agent-driven design systems, CI as a safety net, builders shipping into the codebase.

Great organisation, and a conference designed as an experience. Lots of humour, humanity, lightness and simplicity. A perfect example: not-so-comfortable chairs, introduced with humour, and cushions printed with the Nord Star Product logo for the occasion. My favourite part: the welcome tote bag held a bottle of tripel beer with the Nord Star Product logo.

Lunch, by Barkette.

Happy to see Rémi again, whom I hadn’t seen in ages, behind his food truck, Barkette. Thanks to Xavier Koma and Teddy Lecaillon, whom I was glad to see again, and to the whole organising team.

Very good speakers and talks , and real conversations with some of them, especially Anicet Bart and Roger Ingouacka.

The after party, with Tremolo on the decks.

On video , to get a feel for the atmosphere: Stéphane Cruchon’s recap of the event, and another nice video of the day, by Romain Dao from The Cacatoès Theory.

In short

Nord Star Product Aurora showed teams that really use AI, in very different contexts: a Swedish start-up, a team that merges up to 150 pull requests a day, a Kiabi product team, two CPOs from heritage companies, a founder, a Product Builder. Their shared message: AI speeds up everything that is already solid, and leaves us the choices, the decisions and the fieldwork.

To go further:

The full Aurora edition programme.
Stéphane Cruchon’s video recap of the event, and another video of the day, by Romain Dao.
Jake Knapp’s book Sprint, and Marty Cagan’s Transformed, quoted during the panel.

Were you there? Which talk stayed with you the most?

Sources

Nord Star Product, Aurora edition programme
My notes, the automatic transcripts of the talks and my photos of the day

In the Loop Berlin 2026: key takeaways from the n8n opening keynote

Maxime Lenne — Mon, 19 Oct 2026 15:41:00 +0000

Building a workflow with AI now takes 40 seconds. Checking it takes 40 minutes. That is the line I keep from the
opening keynote of In the Loop, the first n8n conference, on 14 October 2026 in Berlin.

A week ago, I wrote on LinkedIn what I was watching for: MCP, more autonomous yet
controllable agents, human-in-the-loop. I shared my first impressions on the day. Here is
the long version: the figures, Jan Oberhauser’s vision, David Roberts’s product announcements, and what it all means
for people who build with n8n.

In the Loop: n8n comes home

In the Loop 2026 is the first major n8n conference. It took place on 13 and 14 October
2026 at Wilhelm Studios in Berlin, the city where n8n started in 2019. The event website sums it up in one line:
“We’re coming home”.

13 October, Partner Day: invite-only day for partners and ambassadors.
14 October, Ecosystem Day: open to everyone, with keynotes, talks, workshops, an expo and an open mic.
Capacity: 1,200 attendees at most, plus the live stream.
Sponsors: SAP, Accenture, Capgemini, Avanai, Exxeta and SVA.

The opening keynote on 14 October had two parts. Jan Oberhauser, founder and CEO, covered the vision and the “why”.
David Roberts, VP Product & Design, then walked through the product: what has shipped, what is coming and what is still
being explored.

The key announcements

n8n in numbers

Jan Oberhauser opened with the scale reached since the first line of code, written in Berlin in 2019.

Metric	Announced value
Active workflows per month	6.7 million
Fortune 500 companies using n8n	60%
Valuation after the SAP investment	€5.2 billion
Community members	519,000
Published templates	13,000
Integrations	2,300
Ambassadors	340
GitHub stars	207,000 (top 30 of all time)
Active partners	400

The vision: “More capable AI should create more capable people”

Jan asks a single question: as AI gets smarter every month, what happens to people? n8n bets on the path where
humans understand more and stay in control.

His argument is a chain: visibility → understanding → control → agency. Remove the first link and everything
falls. Hence four questions before trusting an automation: what goes in, what happens, what comes out, and who is
accountable?

Two ideas follow:

The UI is the trust layer. It does not fade away with AI; it makes actions, boundaries and responsibility readable.
The more critical the process, the more you need to see it. A personal task: the result is enough. A shared team process: the key steps. A business-critical operation (payroll, customer onboarding, anything regulated): every step.

Visibility follows criticality: the result, the key steps, or every step.

The new positioning: “AI makes intelligence possible. n8n makes intelligence operational.” n8n wants to grow from a
product into an operating system for AI-powered work. Critical work needs four building blocks: agents (judgment
and language), deterministic software (one right answer), automation (repeatable steps at any volume) and people and
teams (ownership and the final call).

A rebrand and n8n Labs

Max, “the original Flowgrammer”, unveiled the new identity. The old logo celebrated the nodes; the new one celebrates
the space between them, where the connection happens. The colour stays.

Jan also introduced n8n Labs, a small team exploring new products under three rules: speed and signal over polish,
build for the capability curve rather than the current model, and default to shutting things down. A first Labs
product was teased at the end of the keynote, with a waiting list.

What will never change

Community Edition: free, self-hosted, with the source open to read.
Model choice: your models, your providers.
Equal treatment of free and paying users.
50% of engineers stay focused on the n8n core.

The product, by David Roberts

David Roberts structured his part around the lifecycle of an automation: build, audit, deploy, monitor, govern.

Already shipped:

Feature	What it does	Announced result
Agents (October)	Agents taken out of workflows and put on the same level, better MCP and Slack support	20% more likely to reach production
AI Assistant	Builds workflows from natural language, on Cloud and self-hosted	1.6M conversations, 800,000 workflows, 8 in 10 users satisfied
MCP for building	Connect Claude, Codex or another coding agent to n8n	3x more workflows, 3.4x more likely to succeed
n8n Connect	Sets up selected credentials automatically	Used by 25% of users, 25% higher success rate

Coming next:

Simpler evaluations. Only 15% of teams run evals, yet teams with rigorous evals have a 70% chance of excellent reliability (32% without). n8n wants to surface only the business decisions and handle the rest with AI.
A new execution engine, rewritten from the ground up: progress saved after every step, a reconciliation suite for stuck jobs, automatic retries only on steps marked safe, sandboxed code, worker pools, resource isolation.
Manifest, a separate observability tool for multi-instance setups: what exists, what runs, who owns it, which systems are touched. In testing, open to multi-instance users.
Governance along four axes: access, guardrails, ownership and quality, security. Eventually with agent identities of their own, narrow, attributable and revocable, instead of borrowing a person’s login.

First look at Manifest, n8n’s multi-instance observability.

What it means for n8n builders

1. Agent first, workflow second. The agent becomes the entry point: you describe an outcome, it finds the path.
But David Roberts is clear: an agent doing the same thing 10,000 times a day is just a very expensive workflow. The rule
of thumb: if the thinking happens once, at design time, it is a workflow; if it happens on every run, it is an agent.

Agent-first, workflow second: agents as the entry point, deterministic logic wherever the work requires it.

2. Make deterministic whatever can be. Reliability multiplies along a chain: 10 steps at 95% each give a system
that is 60% reliable (0.95 to the power of 10). The advice: take business logic out of the agent and wrap it in
deterministic tools. The example given: a single “onboard a user” tool rather than three Slack, Google Workspace and
Salesforce tools left to the agent’s choice.

The agent “figures it out”; the workflow “does exactly what you said, every time”.

3. The bottleneck has moved from building to auditing. You build in 40 seconds and check for 40 minutes. There are
two ways to audit: inspection (reading the logic, the job of the canvas) and testing (running it and checking the
results, the job of evals). Of 1,127 users surveyed, 55% said n8n would lose much of its meaning without the canvas.
The canvas stays.

The new bottleneck: the agent builds in 40 seconds, you check for 40 minutes.

4. Trust is earned in stages. First the agent suggests and the human acts. Then the agent acts and the human
approves spending. Finally the human sets a budget and walks away. Budgets are a partial answer to “approval fatigue”:
clicking “OK” a hundred times without reading.

5. Governance is no longer for large companies only. Soon there will be more agents than employees, even in small
businesses. Shadow IT is no longer a spreadsheet macro; it is a cluster of agents with access to sensitive systems.

Two customer stories show the range. Mercedes-Benz chose n8n as its global automation platform, layered and cloud
agnostic. At the other end, a small US agency turns WhatsApp voice notes into CRM updates: from 4-5 minutes down to
10-20 seconds per note, and a sales cycle cut from 62 to 44 days.

My take: the real battle is visibility in production

What stands out to me is that n8n did not give a “magic AI” keynote. The common thread, from Jan’s talk to David’s
roadmap, is trust: seeing every step, auditing, monitoring, governing. Building with AI has become easy; knowing what
runs on a Sunday night, when nobody is watching, has not.

That is exactly the problem I work on with N8Ninja, my app to manage and monitor n8n automations from Apple devices. I
released version 1.2.1 for the occasion. It answers several of the questions raised on
stage:

What the keynote said	What N8Ninja 1.2.1 does
“The more critical, the more you need to see each step”	Debug logs with the JSON input and output of every node
Monitoring: keeping an eye on what runs	Status, duration and errors of each execution; workflow health over 48 h and 7 days
Trusting it runs when nobody is watching	Failure alerts, with background notifications on Mac
Security, the fourth pillar of governance	Security audit with a risk estimate

N8Ninja runs on iPhone, iPad, Mac, Apple Watch and Apple TV, with n8n Cloud as well as self-hosted instances. Manifest
targets observability for large multi-instance organisations; N8Ninja targets the builder who wants the state of their
workflows in their pocket.

In short

n8n set a clear course in Berlin: AI builds, people understand and decide. Agents as the entry point, workflows for
reliability, and a big push on auditing, the execution engine, observability and governance. The Community Edition and
model choice remain guaranteed.

To go further:

Watch the full keynote on YouTube.
Try the AI Assistant or connect your coding agent through MCP, if you have not yet.
Get in touch with n8n to test Manifest if you run several instances.
Monitor your workflows from your iPhone, Mac or Apple Watch with N8Ninja.

Which announcement matters most to you? Tell me in the comments.

Sources

In the Loop 2026, official website
Livestream: Opening keynote and Product direction (YouTube)
n8n on LinkedIn: the day in pictures
n8n on LinkedIn: community moments
My notes, the keynote transcript and my captures of the live stream

Originally published on the N8Ninja blog.