Sign in

Pelayo Arbués

@pelayoarbues.com
6.7K followers 1.2K following 780 posts

Head of Data Science and AI at idealista.com | Film photographer | Ph.D. in Economics | Sharing my digital garden at pelayoarbues.com | Working full remote from the North of Spain

PostsRepliesMedia
Pelayo Arbués @pelayoarbues.com · 29/01/2026
When you search for a home, it’s rarely “the whole city.” It’s more like: this neighborhood yes, this one no, that other one too, even if it’s far away. With @idealista.bsky.social new multi-area search, you can mix, exclude and add zones in one map search.
idealista map showing the new feature
020
Pelayo Arbués @pelayoarbues.com · 08/10/2025
Just finished reading 3 Millones de Viviendas by @jorgegalindo.bsky.social. One of the sharpest takes on Spain’s housing crisis I’ve seen. Data, history, and a tough question: what kind of country do we want to build? www.pelayoarbues.com/notes/Tres-M... #realestate
Photography of the book Tres Millones de viviendas displayed on top of orange bricks
040
Pelayo Arbués @pelayoarbues.com · 23/09/2025
Context engineering is the new AI buzzword. And here we go again: www.pelayoarbues.com/notes/Contex...
A text editor in baroque style, created using magnific.ai
120
Pelayo Arbués @pelayoarbues.com · 20/05/2025
500 days reviewing highlights on @readwise.bsky.social!!!
251
Pelayo Arbués @pelayoarbues.com · 20/05/2025
I just shared a learning path guide for software developers looking to infuse AI into their products, tailored specifically to your needs, interests, and how deep you'd like to dive. www.pelayoarbues.com/notes/AI-Lea...
A digital landscape showing gradient descent
020
Pelayo Arbués @pelayoarbues.com · 12/04/2025
When we don’t know someone well, we tend to project intentions onto them, casting them as a character in a play they themselves are unaware exists.
Evolution from character to person. When we don’t know someone well, we tend to project intentions onto them, casting them as a character in a play they themselves are unaware exists. It’s quite challenging to optimize on behalf of characters in your mental play, but it’s much easier to be understanding of people you know personally. Spending time together learning about each other, often at a team offsite, will slowly transform strangers into people.
040
Pelayo Arbués @pelayoarbues.com · 01/04/2025
Why software engineers should learn a bit of data science www.pelayoarbues.com/notes/Why-So... #databs
AI generated image of a bridge linking two lands
030
Pelayo Arbués @pelayoarbues.com · 01/04/2025
In simple terms, the purpose of data is to give you a causal model of your business in your head. If reality changes, the figures in your business should show it. At which point you get to update your knowledge.
In simple terms, the purpose of data is to give you a causal model of your business in your head. If reality changes, the figures in your business should show it. At which point you get to update your knowledge.
030
Pelayo Arbués @pelayoarbues.com · 25/03/2025
Coaching is no longer a specialty; you cannot be a good manager without being a good coach. The path to success in a fast- moving, highly competitive, technology- driven business world is to form high- performing teams and give them the resources and freedom to do great things.
Coaching is no longer a specialty; you cannot be a good manager without being a good coach. The path to success in a fast- moving, highly competitive, technology- driven business world is to form high- performing teams and give them the resources and freedom to do great things.
021
Pelayo Arbués @pelayoarbues.com · 24/03/2025
When leaders take steps to protect their own interests, particularly when those steps are taken at the expense of others, they send a message to everyone else that it is okay to do the same.
When leaders take steps to protect their own interests, particularly when those steps are taken at the expense of others, they send a message to everyone else that it is okay to do the same.
011
Pelayo Arbués @pelayoarbues.com · 17/03/2025
To maintain the core of your product there are usually one or two things that have to stay still while everything else spins and changes around them. And that’s a useful constraint. You need some constraints to force you to dig deep and get creative.
To maintain the core of your product there are usually one or two things that have to stay still while everything else spins and changes around them. And that’s a useful constraint. You need some constraints to force you to dig deep and get creative, to push envelopes you hadn’t thought to open before
040
Pelayo Arbués @pelayoarbues.com · 15/03/2025
Yes, leadership is about vision. But leadership is equally about creating a climate where the truth is heard and the brutal facts confronted
Yes, leadership is about vision. But leadership is equally about creating a climate where the truth is heard and the brutal facts confronted
031
Pelayo Arbués @pelayoarbues.com · 14/03/2025
If people only comply, we can’t expect people to take responsibility for their actions. The chain of command is for orders, not information. Responsibility is not doing as we are told, that’s obedience. Responsibility is doing what is right.
If people only comply, we can’t expect people to take responsibility for their actions. The chain of command is for orders, not information. Responsibility is not doing as we are told, that’s obedience. Responsibility is doing what is right.
001
Pelayo Arbués @pelayoarbues.com · 13/03/2025
The push for commercial control of AI capability is dangerous. Naomi Klein, who coined the term “shock doctrine” as “the brutal tactic of using the public’s disorientation following a collective shock… to push through radical pro-corporate measures”
The push for commercial control of AI capability is dangerous. Naomi Klein, who coined the term “shock doctrine” as “the brutal tactic of using the public’s disorientation following a collective shock… to push through radical pro-corporate measures”, is now warning that AI is “likely to become a fearsome tool of further dispossession and despoilation”.
031
Pelayo Arbués @pelayoarbues.com · 12/03/2025
Having two different pipelines to process your data is a common cause for bugs in ML production. One cause for bugs is when the changes in one pipeline aren’t correctly replicated in the other, leading to two pipelines extracting two different sets of features.
Having two different pipelines to process your data is a common cause for bugs in ML production. One cause for bugs is when the changes in one pipeline aren’t correctly replicated in the other, leading to two pipelines extracting two different sets of features. This is especially common if the two pipelines are maintained by two different teams, such as the ML team maintains the batch pipeline for training while the deployment team maintains the stream pipeline for inference
030
Pelayo Arbués @pelayoarbues.com · 09/03/2025
Every three to six months, audit your meetings. Are they useful? Are the right people involved? Do they achieve their purpose, and are they worth the time? Could more be accomplished asynchronously? Consider polling participants or having a neutral person do so.
Every three to six months, audit your meetings. Are they useful? Are the right people involved? Do they achieve their purpose, and are they worth the time? Could more be accomplished asynchronously? Consider polling participants or having a neutral person do so.
020
Pelayo Arbués @pelayoarbues.com · 03/03/2025
Not finance. Not strategy. Not technology. It is teamwork that remains the ultimate competitive advantage, both because it is so powerful and so rare.
Not finance. Not strategy. Not technology. It is teamwork that remains the ultimate competitive advantage, both because it is so powerful and so rare.
061
Pelayo Arbués @pelayoarbues.com · 03/03/2025
Make sure your candidates know that you’re excited about them. I first encountered this idea reading Rands’s “Wanted” article, 4 and he does an excellent job of covering it there. The remarkable thing is how few companies and teams do this intentionally.
Make sure your candidates know that you’re excited about them. I first encountered this idea reading Rands’s “Wanted” article, 4 and he does an excellent job of covering it there. The remarkable thing is how few companies and teams do this intentionally: in my last interviewing process, three of the companies I spoke with expressed interest exceptionally well, and those three companies ended up being the ones I engaged with seriously.
000
Pelayo Arbués @pelayoarbues.com · 24/02/2025
Self-awareness has three components: understanding your underlying value system, identifying your innate preferences—your work style and decision-making tendencies—and being clear about your own skills and capability gaps.
Self-awareness has three components: understanding your underlying value system, identifying your innate preferences—your work style and decision-making tendencies—and being clear about your own skills and capability gaps.
040
Pelayo Arbués @pelayoarbues.com · 23/02/2025
To write something, you have to do things, collect ideas, process them, put words on the page, structure them, edit them, rewrite them, and publish them to your site — and that’s all for one article. Then, there’s the emotional challenges: writer’s block, imposter syndrome (...)
To write something, you have to do things, collect ideas, process them, put words on the page, structure them, edit them, rewrite them, and publish them to your site — and that’s all for one article. Then, there’s the emotional challenges: writer’s block, imposter syndrome, lack of time, judgment from strangers, and criticism from friends.
020
Pelayo Arbués @pelayoarbues.com · 19/02/2025
"Any given executive is almost always uncannily good at one way of consuming information. They feel most comfortable consuming data in that particular way, and the communication systems surrounding them are optimized to communicate with them in that one way." By @lethain.com
Any given executive is almost always uncannily good at one way of consuming information. They feel most comfortable consuming data in that particular way, and the communication systems surrounding them are optimized to communicate with them in that one way.
030
Pelayo Arbués @pelayoarbues.com · 18/02/2025
I would argue operational clarity is exactly the opportunity many data teams have been looking for. So many leaders we speak to say something like “We can do so much more.” Adding operational clarity is something undeniably valuable that data folks are uniquely positioned to do well.
I would argue operational clarity is exactly the opportunity many data teams have been looking for. So many leaders we speak to say something like “We can do so much more.” Adding operational clarity is something undeniably valuable that data folks are uniquely positioned to do well.
020
Pelayo Arbués @pelayoarbues.com · 14/02/2025
Better products, services and experiences are usually the result of the employees who invented, innovated or supplied them. As soon as people are put second on the priority list, differentiation gives way to commoditization. And when that happens, innovation declines (...)
and the pressure to compete on things like price, and other short- term strategies, goes up.
010
Pelayo Arbués @pelayoarbues.com · 12/02/2025
I once read somewhere on the internet: deploying is easy if you ignore all the hard parts
I once read somewhere on the internet: deploying is easy if you ignore all the hard parts
020
Pelayo Arbués @pelayoarbues.com · 12/02/2025
Being able to set ego aside and be ruthlessly critical of yourself and your division is hard. It’s not just about where you have gaps today but also about what you’ll need in the next three to five years.
Being able to set ego aside and be ruthlessly critical of yourself and your division is hard. It’s not just about where you have gaps today but also about what you’ll need in the next three to five years.
110
Pelayo Arbués @pelayoarbues.com · 08/02/2025
"You can’t manage what you can’t measure” is a maxim that is taught and believed by many in both the business and education sectors. But in fact, the phrase is ridiculous. A large portion of what we manage can’t be measured, and not realizing this has unintended consequences
Screenshot of the quote and cover of book
2103
Pelayo Arbués @pelayoarbues.com · 07/02/2025
Just tested Mistral's Le Chat, works really well and super fast. This prompt did a great job: chat.mistral.ai/chat.
Screenshot of prompt: Act like a very experienced Chief Data Scientist. Design a system for tackling fraud in an online marketplace similar to Zoopla.
010
Pelayo Arbués @pelayoarbues.com · 04/02/2025
Facing difficult problems starts by feeling incompetent. It happened to me with a recommendation system during my time as a consultant. This week I share how I managed the pressure through reading, introspection, and the help of my teammates www.pelayoarbues.com/notes/A-reco... #databs
An illustration made with AI depicting a beast in a futuristic library
040
Pelayo Arbués @pelayoarbues.com · 01/02/2025
We think we can easily see into the hearts of others based on the flimsiest of clues. We jump at the chance to judge strangers. We would never do that to ourselves, of course. We are nuanced and complex and enigmatic. But the stranger is easy.
We think we can easily see into the hearts of others based on the flimsiest of clues. We jump at the chance to judge strangers. We would never do that to ourselves, of course. We are nuanced and complex and enigmatic. But the stranger is easy.
030
Pelayo Arbués @pelayoarbues.com · 29/01/2025
Staying late or arriving early or staying home to work in peace is a damning indictment of the office environment.
030
Pelayo Arbués @pelayoarbues.com · 28/01/2025
In ML, weak learners like decision trees have been incredibly effective when combined. Algorithms like Random Forest and XGBoost turn these weak learners into strong models. What if we applied this same principle to LLMs?
Robots pictured in a decision tree
231
Pelayo Arbués @pelayoarbues.com · 22/01/2025
Stakeholders will sometimes pressure you to find data that supports a narrative they have already created in advance. While playing along with this might score you some points in the near term, what will help you in the long term is being a truth seeker.
Stakeholders will sometimes pressure you to find data that supports a narrative they have already created in advance. While playing along with this might score you some points in the near term, what will help you in the long term is being a truth seeker and promoting the narrative that the data truly supports
130
Pelayo Arbués @pelayoarbues.com · 19/01/2025
Instead of condemning people, let’s try to understand them. Let’s try to figure out why they do what they do. That’s a lot more profitable and intriguing than criticism; and it breeds sympathy, tolerance and kindness. “To know all is to forgive all.”
Instead of condemning people, let’s try to understand them. Let’s try to figure out why they do what they do. That’s a lot more profitable and intriguing than criticism; and it breeds sympathy, tolerance and kindness. “To know all is to forgive all.”
120
Pelayo Arbués @pelayoarbues.com · 15/01/2025
Companies that bubble- wrap their employees with tons of free perks are usually shortsighted and have no long- term strategy to sustain those perks, or they have an innately problematic core business and the perks are the cover.
Companies that bubble- wrap their employees with tons of free perks are usually shortsighted and have no long- term strategy to sustain those perks, or they have an innately problematic core business and the perks are the cover.
020
Pelayo Arbués @pelayoarbues.com · 10/01/2025
Give a good idea to a mediocre team, and they will screw it up. Give a mediocre idea to a great team, and they will either fix it or come up with something better. If you get the team right, chances are that they’ll get the ideas right.
Give a good idea to a mediocre team, and they will screw it up. Give a mediocre idea to a great team, and they will either fix it or come up with something better. If you get the team right, chances are that they’ll get the ideas right.
0111
Pelayo Arbués @pelayoarbues.com · 02/01/2025
We stopped talking about building teams, and talked instead of growing them. The agricultural image seemed right. Agriculture isn’t entirely controllable. You enrich the soil, you plant seeds, you water according to the latest theory, and you hold your breath. You might get a crop; you might not.
We stopped talking about building teams, and talked instead of growing them. The agricultural image seemed right. Agriculture isn’t entirely controllable. You enrich the soil, you plant seeds, you water according to the latest theory, and you hold your breath. You just might get a crop; you might not.
080
Pelayo Arbués @pelayoarbues.com · 26/12/2024
Culture is powerful for 3 primary reasons: 1. Because individuals are selected and indoctrinated so well. 2. Because the culture exerts itself through the actions of 100s/1000s of people. 3. Because all of this happens without much conscious intent and thus is difficult to challenge or discuss
060
Pelayo Arbués @pelayoarbues.com · 18/12/2024
I am mentally exhausted. Does anyone else feel like December just wipes them out? www.pelayoarbues.com/notes/Decemb... #dataBS #management
5141
Pelayo Arbués @pelayoarbues.com · 17/12/2024
You must maintain unwavering faith that you can and will prevail in the end, regardless of the difficulties, AND at the same time have the discipline to confront the most brutal facts of your current reality, whatever they might be.
You must maintain unwavering faith that you can and will prevail in the end, regardless of the difficulties, AND at the same time have the discipline to confront the most brutal facts of your current reality, whatever they might be.
130
Pelayo Arbués @pelayoarbues.com · 16/12/2024
“Don’t assume the folks who got you here will get you there.” Be constantly vigilant in answering this question: Do I have the talent I need now and for the next two to four years for my team or company to be a success?
“Don’t assume the folks who got you here will get you there.” Be constantly vigilant in answering this question: Do I have the talent I need now and for the next two to four years for my team or company to be a success?
020
Pelayo Arbués @pelayoarbues.com · 14/12/2024
"It’s poison to think great ideas can only come from you. That you alone can hoard them in one place. And it’s stupid. Wasteful. A CEO has to recognize fantastic ideas— regardless of their source."
It’s poison to think great ideas can only come from you. That you alone can hoard them in one place. And it’s stupid. Wasteful. A CEO has to recognize fantastic ideas— regardless of their source.
140
Pelayo Arbués @pelayoarbues.com · 12/12/2024
"A management technique he called the “rule of two.” He would get the two people most closely involved in the decision to gather more information and work together on the best solution, and usually they would come back a week or two later having decided together on the best course of action."
management technique he called the “rule of two.” He would get the two people most closely involved in the decision to gather more information and work together on the best solution, and usually they would come back a week or two later having decided together on the best course of action.
050
Pelayo Arbués @pelayoarbues.com · 10/12/2024
I just finished reading the Manual of Design Fiction by my friend Fabien Girardin, and I loved it! Design Fiction is a fascinating and powerful technique, especially for those of us working in innovation
The book A manual of design fiction and a notebook on it
240
Pelayo Arbués @pelayoarbues.com · 08/12/2024
Why do I shoot #filmphotography? The colours, the looks.
A stranded bottle on a Galician beach. 

Fiesta devla Jira de Navia, Asturias Views from la Estaca de BaresA street corner in an old village in Madrid.
1150
Pelayo Arbués @pelayoarbues.com · 07/12/2024
"The propensity to lead without being given the authority to do so is what, in organizations, distinguishes people that can innovate and break free of the constraints that limit their competitors. Innovation is all about leadership, and leadership is all about innovation."
Quote: "The propensity to lead without being given the authority to do so is what, in organizations, distinguishes people that can innovate and break free of the constraints that limit their competitors. Innovation is all about leadership, and leadership is all about innovation. The rarity of the one is a direct result of the rarity of the other."
010
Pelayo Arbués @pelayoarbues.com · 07/12/2024
"Because ML systems are part code, part data, and data can change quickly, ML systems need to be able to evolve quickly. This is tightly linked to maintainability." via @chiphuyen.bsky.social
Quote from book: "Because ML systems are part code, part data, and data can change quickly, ML systems need to be able to evolve quickly. This is tightly linked to maintainability."
1100
Pelayo Arbués @pelayoarbues.com · 05/12/2024
Qué bien comunica @sailormerqury.bsky.social y qué buenos mensajes lanza sobre la adopción de la IA al público general. Se necesita más gente como ella divulgando con sentido!
Nerea Luis speaking about AI on a stage
012
Pelayo Arbués @pelayoarbues.com · 04/12/2024
Asturian cider granted Intangible Cultural Heritage by UNESCO. In the picture, my wife's cousin pouring the cider in the traditional way. ich.unesco.org/en/files-202...
A woman pouring Asturian apple cider in the traditional style.
070
Pelayo Arbués @pelayoarbues.com · 03/12/2024
"What you’re building never matters as much as who you’re building it with." - Tony Fadell
Quite from the book Build: What you’re building never matters as much as who you’re building it with.
091
Pelayo Arbués @pelayoarbues.com · 02/12/2024
A clear blue sky over Barcelona highlights a 60-meter-tall column topped with a bronze statue of Christopher Columbus. Erected in 1888 for the Universal Exhibition, this landmark stands at the end of La Rambla, near the city's port. Shot on Kodak Portra 400 with a Nikon FM3 camera and 50 mm lens.
A clear blue sky over Barcelona highlights a 60-meter-tall column topped with a bronze statue of Christopher Columbus.
030