Issue #8 of The Corporate Layer — January 2026

They say that in racing, you either have “the pace” or you don’t. But in the age of Data Science, speed isn’t just a talent, it’s a calculated output.

This weekend, somewhere between the smell of gasoline and the roar of the engines, I didn’t just go to the karting track to drive. I went to build.

I didn’t rely on “gut feeling” or a flimsy paper receipt; instead, I pulled telemetry out of a weekend hobby to uncover the seconds I was leaving on the asphalt.

Here is the story of how I fired the driver (me) to hire a Race Engineer (also me), and built the Garcelán Racing Team.

Here’s how!

🏁 10 Years on the Track, 1 Weekend in the Cloud

I have been karting for over 10 years.

From the moment I was tall enough to reach the pedals, I’ve been chasing that specific feeling you get when you step onto the track. The smell of gasoline, the vibration of the engine, the adrenaline of the first corner.

For 30 minutes, you feel like Fernando Alonso.

But for the last decade, the ritual has always ended the same way. You pull into the pit lane, take off your helmet, and walk to the counter.

Back in the day, they used to hand you a flimsy thermal receipt. (I still keep a few fading ones in a drawer somewhere).

• Lap 4: 57.5s

• Lap 5: 56.7s

Nowadays, it’s a bit more modern. Most tracks have apps that show you the classification: 1st, 2nd, 3rd. But here is the problem:

Knowing I was “3rd” doesn’t help me become “1st”.

These apps are just digital receipts. They tell me the result, but they don’t tell me the process.

• Did I brake too early on Turn 3?

• Did I lose traction on the hairpin?

• Was I fast because of skill, or because I had a lighter kart?

Most people look at the time, say “nice,” and go home. But as I stared at those basic numbers, my “Data Engineer brain” started itching.

I realized I was generating valuable data every millisecond, but it was evaporating. I realized that if I wanted to improve, I couldn’t just be a driver. I needed to be a Race Engineer.

So, I did what I do best: I over-engineered a solution.

🛑 The “Black Box” of Karting

In professional racing, decisions are made on data. In Formula 1, there are hundreds of sensors sending real-time telemetry to a wall of engineers. In amateur karting, decisions are made on “vibes.”

• “I felt fast that lap.”

• “I think I braked late.”

The "8-year-old driver" in me was happy just to drive. But the Data Engineer in me was furious. As an AI Researcher and Data Scientist, "I think" is not a valid metric. I wanted observability.

I have my own helmet and gear, but I was missing the digital layer. I needed to bridge the gap between "driving" and "analyzing."

I decided to build a Full-Stack Telemetry Pipeline. So,

I fired the driver (me) and hired a Race Engineer (also me).

🎨 Over-Engineering a Hobby (The Corporate Layer Way)

Before writing a single line of code for the telemetry app, I had to solve the most important part: The Identity.

If I was going to treat this like an F1 operation, I needed a team name. I couldn’t just call the project “Jorge’s Kart App.” That sounds like a homework assignment.

Introducing: Garcelán Racing Team.

I designed this logo to capture the two worlds I live in:

1. The Nostalgia: The pixel art style represents the arcade games of my childhood (like Mario Kart!).

2. The Reality: The helmet is a digitization of my actual helmet that I wear on the track.

By giving the project a name and a logo, I stopped treating it like “some code I’m writing” and started treating it like a Product for my racing team.

🛠️ The Architecture: From Asphalt to Cloud

I decided to build a Full-Stack Telemetry Pipeline. I used RaceChrono to capture the raw data, but I needed a custom “Mission Control” to actually make sense of it.

Here is the architecture driving the Garcelán Racing Team:

Tier 1: Data Ingestion (The Sensors) 📡

• The Tool: I use RaceChrono to capture high-frequency GPS and accelerometer data during the session.

• The Challenge: RaceChrono is great for recording, but the analysis is stuck on a small phone screen. I needed to extract that raw CSV/JSON data and ingest it into my own system.

Tier 2: The Backend (The Engine) ⚙️

• Processing: Python scripts to clean the noise and calculate sector deltas.

• Logic: I wrote algorithms to calculate my “Theoretical Best Lap” (combining my best sectors from different laps).

Tier 3: The Frontend (The Dashboard) 📊

• The UX: A custom Web App.

• Heatmaps: Visualizing braking zones (Red) vs. Acceleration zones (Green) overlaid on the track map.

• The “Why”: So I can look at a chart and say, “Ah, I’m losing 0.2 seconds because I’m not carrying enough exit speed on the chicane,” instead of just guessing

✨ Why This Matters

Why go through all this effort for a weekend hobby?

Because Applied AI isn’t just about optimizing business revenue or automating emails. It’s about optimizing life.

When you apply professional skills to personal passions, two things happen:

1. You learn faster. I am learning more about real-time data visualization building this kart app than I ever would doing a generic tutorial.

2. You build a story. “I like go-karts” is a hobby. “I built a custom telemetry web-app to analyze my racing line” is a Portfolio Piece.

The Garcelán Racing Team might only have one driver and one engineer (both me), but digitally, it operates like a constructor.

📩 The Final Act: Why Subscribe?

This newsletter is home to #TheCorporateLayer.

The motto is simple: Don’t wait for experience. Create it.

We are often told that we need years of tenure to be effective, creative, or “senior.” But as I showed at the dinner table, the right tools, combined with curiosity, can act as a time machine. They allow us to bridge the gap between “student” and “professional” faster than ever before.

In 2026, I am documenting that exact transition. From the final code blocks of my Master’s Thesis (TFM) at UC3M to navigating the real professional world.

Join me if you want to explore:

• The Tech: Practical AI use cases (beyond just memes).

• The Journey: Real-time updates on breaking into the industry.

• The Mindset: How to leverage technology to punch above your weight class.

If you are ready to build the layer, hit subscribe.

Subscribe now

#CorporateLayer
Building experience before you have it.

Issue #7 of The Corporate Layer — January 2026

What a way to start the New Year. My head is still ringing from the grapes, I’m still digesting the holiday feasts, and my phone lights up with the most dreaded notification in modern history:

"iCloud Storage Full. Upgrade?"

Most people just pay the 9.99€/month "lazy tax". But this time was different. Last summer, I worked as a Solutions Architect Intern at AWS. I spent some time learning about S3 Lifecycles, Intelligent Tiering, and optimizing cloud spend for clients.

When I looked at my phone this Christmas, I didn’t see a “memory gallery.” I saw a poorly architected data lake 🫣. I was paying “S3 Standard pricing” (instant access) for 8 years of data I hadn’t queried since 2015.

I decided to initiate a massive Historical Data Migration Project. Here’s how:

The Inspiration: Legacy Infrastructure (Grandparents)

During the holidays, I spent time at my grandparents’ house, digging through heavy, dusty photo albums. The UX was incredible. No loading screens, zero latency, total immersion 🤣.

I realized: My grandparents were the original Data Archiving Engineers. They understood that "Cold Data" belongs on physical media, not in the cloud.

☁️➡️🏠 The Architecture: Designing a "Hybrid" Strategy

I decided to migrate photos from every trip I’ve taken since 2015 (Exchange year, Summer trips, Weekend getaways) into physical “Deep Archive” albums, architecting a solution based on Access Patterns. Continuing the analogy:

• Tier 1: S3 Standard (My phone) 🔥

  • Data: Memes, photos from this week, QR codes.

  • Requirement: Instant access, high throughput.

  • Action: Keep on device.

• Tier 2: S3 Glacier / Deep Archive (My Bookshelf) 🧊

  • Data: The 50 best photos of the year. The core memories (Trips from 2015).

  • Requirement: 99.999999999% Durability. Low cost. Rare access.

  • Trade-off: High Latency. (Retrieval time: 30 seconds to walk to the shelf).

  • Action: “Migrate” to a Physical Album.

But I had a bottleneck: The Frontend Design (of course, you know me well…). I didn’t want 20 generic albums. I wanted each cover to represent the soul of the location. As a loyal Canva power-user, I have the tools, but I needed a scalable creative engine.

🤖 The Migration Pipeline (ETL with Generative AI)

You can't just "dump" data into a book. That’s bad data engineering. I built a custom ETL (Extract, Transform, Load) pipeline.

1. Extract: I pulled 8+ years of travel data (5,000+ photos) from the Cloud.

2. Transform (The “Gem”):

   • Data Cleaning: I used generic AI to filter out duplicates and screenshots.

   • Enrichment (The Secret Sauce): I didn’t want generic covers. I created a custom Gem (‘The Travel Album Illustrator’) with a specific system instruction:

     “You are a textile historian. Your goal is to generate image prompts for photo album covers based on traditional embroidery and fabric patterns from specific regions.”

     (💡 Psst... If you are interested in the specific system prompt code, drop a comment or send me a DM, and it will be your special Three Kings gift 🎁)

   • The Prompt Strategy: I fed the agent the trip location and context of the trip, and forced it to generate visual concepts based on local embroidery and characteristic textiles.

     “I spent an academic year (August 2022 to June 2023) living in Atlanta as an international exchange student at Georgia Tech.”

   • The Inference: I used the Google’s Nanobanana Pro model to generate hyper-realistic textile textures.

     The representation of Georgia Tech as a fabric quilt.

3. Load: Assembling the final artifacts in Canva (my front-end of choice) and sending to print (Production).

🎨 The Artifacts (The Deliverables)

I took these AI-generated textures and imported them into Canva, adding typography that matches the “Corporate Layer” aesthetic: clean, bold, and intentional.

I treated the "Load" phase like a branding project. Using Canva as my frontend, I merged these hyper-realistic AI textures with the "Corporate Layer" mindset.

The result? Each book looks less like a photo album and more like a coffee table book published by a boutique design agency.

✨ The Insight

During my AWS internship, I learned that being a Solutions Architect isn’t just about knowing the tools; it’s about selecting the right tool for the right access pattern.

• Hot Data (Memes/QR Codes): Stay on iPhone (S3 Standard).

• Cold Data (Life Memories): Move to Album (Glacier Deep Archive).

The technology (Gemini/Canva) didn’t replace the physical experience; it upgraded it.

📩 The Final Act: Why Subscribe?

This newsletter is home to #TheCorporateLayer.

The motto is simple: Don’t wait for experience. Create it.

We are often told that we need years of tenure to be effective, creative, or “senior.” But as I showed at the dinner table, the right tools, combined with curiosity, can act as a time machine. They allow us to bridge the gap between “student” and “professional” faster than ever before.

In 2026, I am documenting that exact transition. From the final code blocks of my Master’s Thesis (TFM) at UC3M to navigating the real professional world.

Join me if you want to explore:

• The Tech: Practical AI use cases (beyond just memes).

• The Journey: Real-time updates on breaking into the industry.

• The Mindset: How to leverage technology to punch above your weight class.

If you are ready to build the layer, hit subscribe.

Subscribe now

#CorporateLayer
Building experience before you have it.

Issue #6 of The Corporate Layer — January 2026

They say a magician never reveals their secrets. But in the age of Generative AI, the “magic” belongs to anyone curious enough to ask the right questions.

This New Year’s Eve (“Nochevieja” in spanish), somewhere between the twelve grapes (yes, it’s a typical spanish tradition from 1909) and the champagne toast, I didn’t just attend the party, I became the entertainment.

I didn’t pull a rabbit out of a hat; instead, I pulled reality-looking pictures out of my phone.

Here is the story of how I finished my year by turning into a digital wizard, and how I plan to start the new one (btw, Happy New Year 🥳).

🎓 The Origin Story: The Apprentice

To understand the magic, you have to understand the training.

I’ve always been the type to tinker with technology, fascinated by how lines of code and data can shape reality. That lifelong curiosity eventually led me to the deep end: the Master in Applied Artificial Intelligence at UC3M (Universidad Carlos III de Madrid).

I have officially finished the coursework. The lectures, the complex implementation challenges, and the late-night debugging sessions are finally behind me. All that stands between me and the finish line now is the final “magic trick”: TFM (Trabajo Fin de Máster / Master’s Thesis).

Spending a year immersed in neural networks and machine learning changes your perspective. You stop seeing “magic” and start seeing probability, weights, and biases. But when I went home for the holidays,

I realized that to my family, what I do is still pure sorcery.

🍇 The Performance: “Nochevieja”

The scene: A loud dinner table, empty plates of jamón, and the inevitable lull in conversation before the countdown.

I decided to wake everyone up. I pulled out my phone (my wand) and opened Gemini and ChatGPT.

My goal? To take our ordinary family context and turn it into something absurdly funny using AI (multimodal AI if we are technical…).

I wanted to roast our traditions and create “edits” of our reality.

The Spell (The so called “Prompt”)

I started snapping fresh photos across the dinner table and digging up old gems from the family WhatsApp group and the old physical family albums.

The “magic” was in asking the AI to keep our faces perfectly recognizable while completely rewriting the context and wardrobe.

I ran two types of prompts:

The Prank (Hyper-Realism): “Take this photo of my grandfather. Place him on the sidelines of the Santiago Bernabéu stadium, dressed as the Real Madrid coach giving orders. Make it look 100% authentic, like a press photo.”

The Comedy (Style Transfer): “Take this group photo and reimagine us as a traditional Spanish family from the 80s (shoulder pads included), then as medieval royalty, and finally as flower-power hippies.”

These are just some examples :)

The Result

The table erupted, but the performance actually had two acts.

• Act 1: The “Is That Real?” Test

  First, we wanted to see if they would bite. I didn’t make them look like cartoons immediately; I aimed for believability. I showed them the image of my grandfather commanding the squad at the Bernabéu, and then another one of him crossing the finish line of the San Silvestre Vallecana (Madrid’s iconic New Year’s Eve 10k), sweating and triumphant. The lighting and texture were so convincing that for a split second, they were genuinely confused. “Wait, when did he go there? Is that old footage?”

• Act 2: The Comedy

  Once the line between reality and AI was blurred, we leaned into the absurdity. That’s when we transformed the whole family into a 1980s sitcom cast, medieval knights holding chicken legs, and grandparents partying on a luxury yacht.

That is the hook: It wasn’t just about making funny pictures; it was about remixing our lives. It turned the dinner table into a playground where my grandfather could be Ancelotti for five minutes, and my grandmother could be a hippie icon 😂.

(Yes, Gemini’s model NanoBanana Pro does a really good job 👀)

🔮 Demystifying the Magic (AI Divulgation)

Here is the reality behind the illusion: it’s just math mimicking language and vision.

When I put that prompt into Gemini or ChatGPT, I’m not casting a spell. I’m guiding a model that has read more text and seen more images than any human could in a million lifetimes.

• Transformers: These are the engines behind the curtain. They pay “attention” to different parts of my request (the “shrimp,” the “Renaissance style”) to predict the best possible output.

• Prompt Engineering: This is the wand movement. The specific words I used (”surrealist,” “roast,” “dramatic”) adjusted the parameters of the generation.

I have been thinking a lot about this idea (and I know that there are some controversies too around it) but the real power of AI isn’t in replacing human creativity, it’s in acting as a multiplier for it.

It allowed me to take a passing thought:“this dinner is a circus”, and visualize it in high fidelity instantly. It wasn’t about replacing art, it was about creating pictures easily and fast, just to spark fun and shared laughs around the table.

✨ The Next Act: 2026

As the clock struck midnight and we ate the grapes (mostly on time), I realized that this is exactly what I want to do with my career.

My Master’s at UC3M gave me the technical foundation. Understanding the how.

But the real joy comes from the what and the why.

It comes from bridging the gap between complex algorithms and a family laughing at a generated photo of a dog in a tuxedo.

What’s next?

1. Crushing the TFM: The final “act” of my Master’s degree.

2. More Divulgation: I want to keep writing about AI, not just as code, but as a creative tool.

3. More Magic: Finding new ways to use these tools to make life a little more efficient and a lot more fun.

Happy New Year. May your prompts be clear and your hallucinations be minimal.🥂

📩 The Final Act: Why Subscribe?

This newsletter is home to #TheCorporateLayer.

The motto is simple: Don’t wait for experience. Create it.

We are often told that we need years of tenure to be effective, creative, or “senior.” But as I showed at the dinner table, the right tools, combined with curiosity, can act as a time machine. They allow us to bridge the gap between “student” and “professional” faster than ever before.

In 2026, I am documenting that exact transition. From the final code blocks of my Master’s Thesis (TFM) at UC3M to navigating the real professional world.

Join me if you want to explore:

• The Tech: Practical AI use cases (beyond just memes).

• The Journey: Real-time updates on breaking into the industry.

• The Mindset: How to leverage technology to punch above your weight class.

If you are ready to build the layer, hit subscribe.

Subscribe now

#CorporateLayer
Building experience before you have it.

Issue #5 of The Corporate Layer — December 2025

Every December, our feeds turn into a sea of neon greens and vibrant pinks.

It’s Spotify Wrapped season.

We share our top artists and our “audio auras,” celebrating the soundtrack of our year.

But as I scrolled through my own stats, I realized something:

My 2025 wasn’t defined by what I listened to.
It was defined by what I built.

2025 was the year The Corporate Layer went from a personal “hack” to keep myself motivated, to a framework that changed how people perceive me.

I decided that if I wanted to be seen as a professional, I had to stop waiting for a badge and start designing it myself.

💡 The Coffee That Started the Playlist

A few months ago, my supervisor told me something that stopped me cold.

A professor from last year still remembers my work.

Out of hundreds of students.
Thousands of Jupyter notebooks.
Endless identical assignments.

Why me?

Because I never handed in homework.

I handed in products.

I gave assignments a name.
A visual identity.
A context.
A reason to exist.

I added The Corporate Layer.

(That moment is what started this entire Substack! You can read the original post here.)

🚪 Escaping the “Junior Trap”

That coffee conversation was the validation I needed for a fear that had been haunting me all year. At the start of 2025, I felt the “Junior Trap” closing in.

I was at UC3M, surrounded by brilliant peers, and I realized that having the same Master’s degree as everyone else meant I was practically invisible. I was terrified of being “just another resume” in a stack of thousands, a line of text waiting for permission to exist.

The Corporate Layer wasn’t born out of confidence; it was born out of a need to stand out.

I had been quietly applying this “over-designing” mentality to my work for a while, but this year, something shifted. The anxiety of being “just a student” turned into the fuel for a framework.

I realized that experience isn’t something you wait for. It’s something you build.

Looking back at 2025, I didn’t just survive the curriculum; I architected my own reality.

Here is what my year actually looked like, beyond the grades.

📊 My 2025 Stats (The Numbers That Actually Mattered)

If my year were a Wrapped card, the stats would look a little different than a top listener’s:

1 Master’s in Applied AI at UC3M (Almost!):

I’ve survived the formulas, the models, and the sleepless nights. Only the TFM (Final Master Thesis) stands between me and the finish line, but the real “Applied AI” happened in the layers I built around the code.

16 “Fake” Companies Created:

Every assignment was an excuse to play CEO. I spent more time choosing hex codes for non-existent startups than I did debugging code.

1 Stage Conquered at T3chFest:

Standing there, talking about AI and the US Elections, I realized that technical knowledge is useless if you can’t tell a story that makes people lean in.

2,800 Kilometers to Brașov:

Traveling to Romania for womENcourage 2025 to present my work as a researcher at UC3M. It was the moment I realized my research wasn’t just “academic”, it was my contribution to a global conversation.

3 Months at AWS:

My internship as a Solutions Architect was the ultimate stress test. I wasn’t just learning cloud infrastructure, I was learning how to architect value.

My Top Genre: “Academic Rebellion (feat. Imposter Syndrome)”:

The art of over-designing a mundane assignment until it looks like a Fortune 500 pitch, while constantly wondering if I’m taking it too far (Spoiler: I think I’m not).

🏆 The “Artists” I Had on Repeat (My Top Tools)

My 2025 soundtrack wasn’t a playlist; it was the set of principles I leaned on every time I opened my laptop:

1. Storytelling: The “lead singer” of my year. Because a Solutions Architecture diagram at AWS without a story is just a bunch of boxes and arrows.

2. Visual Design: My “bass player.” It’s the foundation that makes people take my AI models seriously before they even run the first cell.

3. The UC3M Research Riffs: Using my role as a researcher not just to find data, but to find insights that look professional enough to be published (and remembered).

4. The AWS Beat: Learning that in the “real world,” the Corporate Layer is actually called “Customer Obsession.”

5. Audacity: The sheer nerve to treat a classroom assignment like a high-stakes pitch for a tech giant.

✨ Looking Ahead to 2026

2025 was about reflecting on the framework.
2026 is about scaling it.

The Corporate Layer isn’t about pretending.

It’s about moving from:

• “doing homework” → building evidence

• “waiting” → designing

• “hoping to be noticed” → being remembered

If you’re waiting for a job offer to start acting like the professional you want to become, you’re missing the point.

Start now.
Add the layer.
Make them remember your name.

What was your “Top Artist” (your biggest win) this year?

And more importantly, what “Corporate Layer” are you adding to your work in 2026?

#CorporateLayer
Building experience before you have it.