Decentralized organizations working like a swarm with AI

You probably know Azeems Azhar’s Exponential View. Also, there’s a podcast, and the last time invited Daniel Hulme, CEO of Satalia, who explained his radical vision on how technology can be honest to build fundamentally decentralized organizations.

I recommend you to listen to it, only 38 minutes, but if you want to know more first, keep reading and you’ll find some insights (spoiler alert).

The Vision

Think about an organization with no managers, no KPI, no holiday control, evaluated by peer review… It seems chaotic, isn’t it? Daniel assures that it works.

The vision is to try to create an organization that works like a swarm, where people have the freedom to work on the things that they want, how they want to work, where, when,… And being able to then remunerate them in a fair way.

This is Daniel’s hypothesis: By being able to operate like a swarm by allowing the talent to move around fluidly and giving them autonomy around the decisions that they are making, they are able to be exceptional at services and products. And as Daniel points out, there are very few companies that are good at both.

Daniel has a long term aspiration that is to scale this to the entire planet. He wants to change the world.

How does it work internally?

They encourage people to look across the available projects and join the ones that they feel attracted to the values. People are not asked to necessarily join in a particular role, they can split effort in a more granular way, joining different projects, and assuming different roles.

The interesting part is that all this work (data) is stored in a platform that allows them to understand how people contribute to the projects and also helps them understand how to contribute to the employee’s career.

They focus on two things: contribution quantity and quality. The first one is easy to capture, it is mainly based on the data stored from the applications used at work. The second is based on peers, like 360 reviews.

Machine Learning also determines who might be the right people to contribute to employee development. They use optimization to maximize the quality of feedback and minimize the quantity.

On top of that, there’s a strong culture and set of principles. Everything they put in place has to be aligned with 4 key principles: equality, freedom, safety, and justice. If any process violates that, it is not implemented.

And what about customers?

Satalia is a company that builds products and services for real customers, for example, retailers like Tesco. They work on large difficult software builds, like a last-mile solution, or a resourcing allocation service.

When a customer asks them to build a product or service, then they socialize that opportunity across the organization, and they also use machine learning to identify who might want to join the project.

The challenge is when somebody wants to leave the project, but as Daniel shared, most people don’t.

Because of the huge transparency across the organization, people are aware of who’s working really hard, doing the best for the organization, and when it comes to peer review and setting salaries, those types of considerations are taking into account by the community.

Challenges and future

The real big challenge is accountability. How could all these employee skills be put in accountability? Is everything accountable? What about leadership? How to bridge between accountability and authority? Responsibility is key when you allow people to move freely with autonomy around decision making.

Also, machine learning doesn’t have to become an element of surveillance. And also access to data has to be controlled. For example, when somebody struggles, they should be able to identify them from a human perspective, to try to help. Machine learning is biased by nature but the question is if it is much or less biased than humans. So the goal is to combine it with humans for decision making.

Another important challenge is growth, and how they could keep culture and principles. This usually happens in all companies with rapid growth. Remember Tony Hsieh’s story about LinkExchange, where they started being friends, and friends of friends, with the same interests and purpose, and ended being strangers with no purpose or commitment.

Regarding the future, also they are facing big questions about their own business model and its competitors. Daniel shared an interesting point of view. “I have a delivery solution that is better than Amazon’s, but how could I compete with Amazon. The answer is you can’t. The only thing you could do is to change the rules of the game. How? make it open to everybody, and it gives you access to a community of talent and data passing through these innovations”. The idea is to decentralize innovations and democratize access to them.


There are many open questions regarding this kind of organization working like a swarm. It is like an agile frameworks evolved to the extreme, with the help of machine learning.

I might admit that the idea of this huge freedom is very cool, but probably it is in a very early stage. But it is great to see companies like Satalia to develop it. It will worth to keep an eye on Daniel’s proposal. Who knows, maybe in the future we will be working as bees.

Did you listen to Azeem’s podcast? Any thoughts about this new kind of work?

I hope you liked it. If so, please share it! Do not hesitate to add your comments. And, if you want to stay up to date, don’t miss my free newsletter.

As always, thanks for reading.

Join the FREE Newsletter

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.