Click Here To Copy The Free Workflow

⚠️ A quick note on security and privacy is included further down.

📢 Disclosure: I work for n8n in their marketing department, but I am not compensated for any of the links below. The opinions expressed here are my own and do not reflect those of n8n GmbH.

Why I Built This

If you’ve ever used YNAB (You Need A Budget) , you already know it’s both brilliant and brutal.

It forces you to approve and categorize every transaction you make. That discipline is what makes YNAB powerful, but it also makes it difficult to be consistent. If you're not consistent all these transactions pile up and coming out from under it is even more difficult as you know how much work is piling up. 

The memo field, by default, is blank. So when transactions import, you’re often left with payee lines like:

“AMZN DIGITAL 1234”
“APPLE.COM/BILL”
“GOOGLE *SERVICES”

None of that tells you what you actually bought. 

Yes, there is a link to the purchases page, but that is yet another tap to open, a scroll to find, and then another tap to close. And then you still have to decide whether to type the information into the memo field for long term storage. 

I’ve spent years trying to improve this, from reverse-engineering Amazons ever-changing email html structure to trying fruitlessly to reverse engineer their backend api. Sometimes workarounds helped for a while, then eventually broke.

Once AI became practical to integrate, the idea finally clicked:

“Why not just feed the entire receipt email to an AI and have it tell me what I bought?”

That realization led to this workflow. In the past I had to know exactly where in the HTML to look to know what was purchased. With AI, even if the HTML structures changed, the AI could still parse out the values inside it. 

Who Is This Workflow For?

If you’re new to automation, deploying an n8n workflow can involve a learning curve. Because of that, this workflow is best suited for more technical users.

The good news is that your cost, in time or money, depends largely on how hands-on you want to be.

Completely Free (Self-Hosted)

This option requires the most technical knowledge, but offers the most control. You’ll need to deploy n8n on your own infrastructure, typically using Docker.

When I say “free,” I mean aside from the electricity running your machine. You’ll need:

• n8n Community Edition (runs free in Docker)
• An SSL solution such as Traefik or another free provider
• A DNS provider like Cloudflare for authentication and tunneling
• An AI provider (LM Studio, Ollama, or Gemini’s free tier)
• The workflow template

If you already have these things in your homelab, you probably already have n8n deployed as well. 

Low Cost Using n8n Cloud

This is the fastest and least technical option. You don’t manage infrastructure, just configure the workflow and credentials.

• n8n Cloud (this workflow runs hourly, which is 730 times a month to give you an idea of cost)
• An AI provider: LM Studio, Ollama, Gemini, or paid services like OpenAI or Anthropic
• The workflow template

How the Workflow Works

This n8n workflow connects YNAB, your email, and an AI model to automatically fill in missing memo details.

The current version focuses on Amazon purchases because Amazon receipts are consistent and made a good starting point. However, the workflow is intentionally designed so you can extend it to Apple, Google, and other merchants with minimal changes.

The workflow runs on a schedule, webhook, or manual trigger. It pulls all unapproved transactions, filters those missing memos, and attempts to match them with receipt emails within a ±5-day window.

When a match is found, the email is sent to an AI agent, which extracts purchased items and prices. The transaction memo is then updated with something readable, such as:

“Amazon – USB-C Hub & HDMI Cable – $32.48”

If no email match is found, the workflow writes a simple fallback memo so the transaction doesn’t get stuck retrying.

In practice, it’s accurate about 80–90% of the time, and it has significantly reduced my weekly budgeting cleanup.

Copy the workflow here:
https://go.n8n.io/amazon-ynab-sync
Setup notes included.

A Note on Security and Privacy

This is where the security-conscious side of me needs to speak up.

When you use a cloud-based AI model, you are sharing detailed purchase history and portions of personal email data with that provider.

That data may be logged, analyzed, or used for training, and you generally have no way to independently verify how it’s handled.

That doesn’t mean it’s unsafe, but it is something worth thinking about. For example if you use Gmail and then use Gemini for AI processing, the data is staying in the same ecosystem. This doesn't mean it's safe, but it does give you more context. 

The Local AI Alternative

While more complex than using a hosted AI service, I’ve had good results running LM Studio locally on my M4 Mac as the AI provider for workflows like this.

For more advanced setups, you can use Ollama , which runs large language models locally on macOS, Linux, or Windows. It’s powerful, but requires more setup.

In my spare time, I’m also experimenting with a fully on-device version of this workflow.

On iOS, Apple exposes an on-device language model via the Shortcuts app, allowing emails to be processed with no cloud AI and no internet connection.

It’s a slower side project, but it’s the direction I’m most excited about.

I’ve already published a related workflow using local AI on iPhone , mainly to show that this approach is feasible.

Hosting Options

I personally self-host n8n using the Community Edition to maintain full control over my data.

If you prefer not to manage infrastructure, n8n Cloud works just as well.

Final Thoughts

This started as a small personal automation, but it’s become a permanent part of my budgeting workflow.

The AI isn’t perfect, but it’s accurate enough to make budgeting feel far more manageable. And with on-device AI becoming practical, I’m excited about what privacy-first automation enables next.

If you use YNAB and shop on Amazon, this workflow will likely save you time. And if you’re comfortable extending it, it’s a solid foundation for Apple, Google, and other receipt-driven services.

(You’ll find the newsletter link in the site footer if you’d like updates on future workflows.)

When you travel for work, it’s easy to fall into the same old rhythm: airports, hotels, events, repeat. But sometimes, a trip sneaks up on you and turns into something far more meaningful than what your calendar promised.

That’s what happened to me in early October 2025 when I traded the heat of Tucson, Arizona for the cool, electric pulse of Berlin, Germany to attend and speak at n8n Builders Berlin — a community event for automation enthusiasts, AI explorers, and makers of every kind.

I went to teach people about workflow automation using n8n. I left with new friendships, late-night stories, and a fresh perspective on why community, not code, is what really powers innovation.

🌍 Why Berlin?

If you’ve never been, Berlin is a fascinating contradiction. It’s a city built on scars, from war, from division, from constant reinvention, yet it thrives on creativity and openness.

It’s industrial, modern, artistic, and raw all at once.

Every corner tells a story. One block might house a centuries-old church, the next an underground art gallery or a startup hub full of hackers and dreamers. It’s a place that constantly tears itself apart just to rebuild better.

So when n8n announced Berlin as the next stop for its Builders event, it made perfect sense. Automation, after all, is about transformation — and no city embodies that better than Berlin. And, it's where n8n is headquartered as well.

🛬 Day 1 – Arriving in Berlin: Reunions and a Live Stream

My flight from Tucson was mercifully uneventful, the best kind of flight when you’ve got work ahead. I landed in Berlin, checked into my hotel, and immediately dropped my bag, showered, and set out to meet the team at n8n headquarters.

Walking through the doors was surreal. Many of these faces: JP, Tom, Niklas, my boss, and others were people I’d collaborated with for a long time but rarely met in person. Screens flatten relationships; handshakes and laughter bring them back to life.

Bart was hosting the n8n Community Stream, and I joined as a guest to talk about the work I’ve been doing. You can even catch that stream here (I show up around the 52-minute mark).

Afterward, the team and a few of our content creator partners went out for dinner. It was one of those nights where the conversation effortlessly drifts from work to life, from automation to travel, from serious to silly. The food was great, the drinks were better, and the company was the best part of all.

⚙️ Day 2 – Work, Collaboration, and the Big Event

The next morning started quietly. Breakfast at the hotel, a few emails, then straight to the office to get some real work done before the big event.

I spent part of the day working with Ian from Puerco TV on our Google Gemini AI collaboration video, and later recorded a session with Mary Newhauser on RAG and agentic workflows.

Mary is one of those rare people whose intelligence feels effortless. Collaborating with her was like taking a crash course in both humility and inspiration. By the time we wrapped, I was buzzing with ideas.

As evening approached, it was time to pack up and head to the venue, Lohmühleninsel, a scenic event space about 30 minutes from the office. The sun was just starting to dip when I arrived, and the place was already alive with energy: banners, gear, people mingling and setting up.

My talk was scheduled for 6:30 PM, so I spent the pre-show hours in the speakers’ lounge, sipping water, tweaking slides, and going over my notes for the Star Wars Demo, a presentation designed to teach beginners how to use n8n in a fun, approachable way.

When I stepped onto the stage, I asked how many people were new to n8n. Nearly 90% of the crowd raised their hands. Perfect. That’s exactly who the demo was for.

The Star Wars Demo uses the Star Wars API to pull movie titles and release dates into a spreadsheet, teaching the fundamentals of automation in a playful way. It’s simple, creative, and approachable, because the truth is, technical learning sticks better when it’s fun.

The crowd was engaged from the first minute. People laughed, asked questions, and by the end, I could see the spark, that moment when something clicks. It’s one of the best feelings in the world.

🤖 The Roundtable: Talking the Future of AI and Automation

Later that evening, I joined four incredible people for a roundtable discussion on the future of AI and automation:

• Max Tkacz, n8n’s original “Flowgrammer”
• Goda Go, a TEDx speaker and content creator with over 100k followers
• David Roberts, VP of Product and Design at n8n
• Marcel Claus-Ahrens, aka Dr. Pure Eval — an automation expert and one of the most entertaining personalities in our community

The discussion was lively, fast-paced, and full of insight. We talked about where automation is heading, how AI is changing workflows, and how human creativity fits into all of it.

There was no script, just a free exchange of ideas among people genuinely passionate about the craft.

When the panel wrapped up, the speakers were gifted Tiny TV sets. Tiny, fully functional televisions from Tiny Circuits. It was such a thoughtful gesture and perfectly symbolic of our shared love for clever engineering.

And then came the networking session. Food, drinks, and deep conversations with content creators, startup founders, automation enthusiasts, and AI experts from around the world.

The vibe felt electric.

The night didn’t end there, it continued at a local bar called QBA, where laughter and music filled the air until around 2 a.m. It was the perfect Berlin ending: spontaneous, spirited, and full of connection.

☕ Day 3 – Improvisation, Coffee, and Unexpected Adventures

The next morning, a big group of us, about twenty of us, decided to cowork together at a shared office space. But Berlin had other plans. It was a public holiday, and the space was closed.

Instead of giving up, we pivoted. A few blocks away stood an old hotel we’d used during a previous Berlin IRL trip. We wandered in, asked if we could use their lobby to work and order coffee, and they graciously agreed. Bart, ever the team player, put coffee for everyone on the company card.

So there we were, laptops open, caffeine flowing, conversations humming softly in the background. It wasn’t the plan, but it worked out perfect.

Later that evening, after dinner, I got a message from Marraliza, an n8n ambassador and fellow creative spirit from Amsterdam. She mentioned her friend from Thailand was visiting Berlin with his girlfriend and invited me out to join them for drinks.

I hesitated for half a second. I was tired, content, ready for bed. But something told me to go.

And I’m glad I did.

That night turned into a full-blown adventure. We met strangers who quickly became friends, wandered through hidden bars, and laughed until our voices were hoarse. It was one of those rare nights that remind you why you travel: for the unexpected human moments.

By 3 a.m., the group had transformed. Some people left, others joined. Conversations about automation gave way to philosophy, jokes, and life stories.

When the night finally ended, it wasn’t the same group that had started it, but that’s the beauty of travel. You meet new people, and for a few fleeting hours, your worlds overlap in perfect harmony.

🌙 Day 4 – Souvenirs, Solitude, and Scooters at Midnight

My body finally called in its debt. I slept through the morning, the afternoon, and most of the day. When I finally woke up, the clock read 5 p.m. I had missed every breakfast and lunch invite, but honestly, it felt earned.

That evening, I took myself out on a solo date with Berlin.

Dinner at Rice & Shine, a cozy spot with flavorful dishes and the comforting hum of quiet conversation.

After dinner, I went hunting for souvenirs — something small to bring home to my sons, Jett and Lucas. I found a candy shop with European treats and grabbed a few giant Kinder Eggs.

Then I wandered into the AMPELMANN Shop Am Tacheles, a cute little store themed around Berlin’s iconic traffic light men, the Ampelmännchen. I left with a few packs of vegan gummy candies shaped like those little walking figures.

Finally, I rented an electric scooter and took off through the empty streets.
Berlin at night is magical. The air was cool, the roads almost deserted, and the city lights reflected off the river like pixels in motion. I flew through intersections, past murals and statues, laughing out loud into the wind. It felt like pure freedom.

I capped the night at a Spanish restaurant called La Esperanza, where I ordered a spread of tapas and watched the city wind down around me.

It was quiet. Peaceful. The kind of solitude that fills you up rather than drains you.

✈️ The Flight Home and What Stayed With Me

The next morning came early. My flight home left at dawn, and as the plane climbed above the city, I looked down at Berlin, a patchwork of memories, lights, and late nights.

This trip wasn’t just another work event. It was a reminder of why I do what I do, not just to build things, but to connect with people who are building their own things too.

n8n Builders Berlin wasn’t just about automation. It was about community, creativity, and human connection in an age of machines.

💭 Final Thoughts

So here’s to the next trip, to more talks, more demos, more conversations over cold drinks and hot workflows.

And maybe, just maybe, another scooter ride through the empty streets of a city that’s always becoming something new.

The trip started like most bad trips do: with weather, delays, and the slow grind of airport bureaucracy. Tucson to Denver was supposed to be a short hop. Instead, we tangoed with a storm that dropped us in Wichita for refueling, then limped back to Denver hours too late. By then, my connection to New York had already gone without me.

The airline gave me a hotel voucher, a meal voucher, the usual sympathy packet of corporate apology. Four hours in a bed that still smelled faintly of the last guy’s deodorant, then back on a plane at dawn. By the time I finally hit New York, I was running on bad sleep and caffeine.

The hotel near Chelsea had a view worth a moment’s pause: the Empire State building looming like a weary old heavyweight, streets pulsing below with the kind of energy only New York manages at 8 a.m. I dropped my bag, inhaled the city grit, and walked to the office.

Inside, chaos and creativity. Max—my colleague—was hunched over a robot that would later be kicked around for a shoot. Cookies materialized from somewhere (New York has a way of producing perfect little indulgences out of thin air), and the fridge was stocked like we were actually going to need hydration for whatever came next.

By late afternoon, we were on a rooftop in the East Village with the CEO. One of those rare, candid moments where leadership pulls back the curtain. A circle of us, tossing questions like darts, him answering with surprising directness. 

Then came the party. My colleague Liam and I presented the new features of n8n to a room of creators, engineers, and flowgrammers. Afterward, the real show began: conversations, questions, the cocktail party current of New York’s tech underground. By the time I made it to the bar, it was already shut down. Maybe that was for the best!

Here is a short video showing the meetup! If you squint you can see me a few times throughout the video!

The next morning, breakfast with Lucas, another colleague, before heading into a recording studio. We interviewed a customer about how they used the platform. No marketing fluff, just real talk about work, automation, and how people bend tools to their will.

Then came the main event: The Agentic Arena. Esports meets reality show, but instead of fragging each other in Halo, contestants built workflows in real time for a shot at ten grand. It went on for a while, 11 a.m. to 6 p.m.,but watching the tension build, the crowd lean in, it was its own kind of spectacle. Afterward: pizza, networking, more conversations. The New York way.

Dinner that night was with creators, the kind of dinner where you leave with more questions than answers, but also a strange sense of community. I met Nate Herk, the king of n8n on YouTube by views. I ended the night with cheesecake from a bakery (souvenir sugar) and a tres leches cake that I probably didn’t need but couldn’t resist.

The flight home was mercifully uneventful. No storms, no diversions, no deodorant-scented hotel layovers.

New York is never easy. It’s a fight every step of the way, against weather, crowds, your own exhaustion. But it’s worth it. Every time.

Success measured by output. We got a lot done. A (mostly) working UI and full database schema ready to go. In that time we spent a lot of time at the Salesforce tower. A really pretty building that we had the priviledge to work there. 

Day 1: Hitting the ground running. 

I landed and pretty quickly went to the office to work with Dylan. On a steady diet of coffee, we pumped out a beautiful website UI and database scheme that worked with n8n to provide a powerful backend, and a useful front end. I can't wait to demo it!

Day 2: Work and Prep for Meetup

It was really difficult to stop working on the Content Creators dashboard and shift gears to prepping for the community meetup, but we did, and it went really well. A lot of the things that were issues before, like having too many speakers were fixed by having less speakers and more Q&A time. I think it really made the audience feel involved.

Day 3 - Work from Hotel Lobby Day

The last day was pretty uneventful. I hung out in hotel lobby and worked on various projects and had a few meetings. 

Also, lesson learned, a USB C fast charger will trickle charge a Macbook Pro. 

I just wrapped up a quick solo work trip to New York City—cool weather, a packed schedule, and just enough downtime to explore the city in my own way.

✈ Getting There (Eventually)

This trip started with a hiccup: what should have been a Wednesday night arrival turned into a bleary-eyed 4 a.m. Thursday check-in, thanks to a United Airlines outage. On the bright side, my Houston leg was in first class, which softened the blow.

🏨 Staying in the Flatiron

I stayed at the Freehand Hotel in the Flatiron District. The place had character—modern design, good location—but did try to sneak in an extra fee. Luckily, my company covered it.

☕ Bagels & Pizza (Because New York)

My first morning, I walked to our NYC office, grabbing a breakfast bagel from Gramercy Bagels along the way. It was exactly what you want from a New York bagel—chewy, fresh, and satisfying.

Dinner highlights included East Village Pizza (classic NYC slice vibes) and Tenjou for ramen, which came with a surprise: the noodles and broth were served separately. It took me a moment to figure out the hidden broth compartment, but once I did, it was pure comfort food.

💼 Work That Didn’t Feel Like Work

The NYC office is less “cubicles and fluorescent lights” and more “incubator coworking space,” which made it a great spot for filming some content. I spent time chatting with my coworkers, especially Eunice and Zarif, and enjoyed swapping stories over pizza one night.

The main reason for the trip was to present at a customer training session. Initially, my section of the demo was going to be cut, but the client ended up giving us the final hour—a slot no other vendor (including Google Firebase or Microsoft Power Automate) received.

During the session, the client champion invited attendees to stick around for a deeper dive with me. Seven people stayed, and three of them were so engaged that what was supposed to be 30 minutes turned into a full hour. I showed three demos:

• Star Wars workflow
• Request Router
• Call Forge for AI prompt management

The LinkedIn request router especially caught their attention, they were already imagining how to repurpose it for email.

🎥 Behind the Scenes

After the demo, I headed back to the office with Liam (a newer teammate who’s sharp and detail-oriented) to film a benchmarking short. Later, I walked back to the hotel in the cool NYC air—a small, refreshing moment in a packed trip.

🛍 Souvenirs & Sights

One highlight was The Friends Experience, where I picked up some souvenirs. I also took some time to just sit on a park bench, with two strangers, and watch the city go by. There was something so peaceful about just sitting there, no place to go or be, and just take in the sounds and sights.

🧠 Reflections

What I loved:

• The energy of presenting and seeing people excited about the demos
• Great food (those bagels…)
• Making progress on big work projects even in between events
• Feeling safe and relaxed walking alone at night

What I didn’t love:

• That unnecessary hotel fee attempt
• That there weren’t more n8n employees to interact with.

Overall, the trip was the perfect mix of productive and restorative. I left feeling inspired, connected to my team, and reminded why I love the energy of New York, especially when I can blend work with little slices of everyday city life.