Coming from web3, an industry started by Cypher punks, I would say the most important fact that AI makers should keep in mind is: privacy matters. AI companies like Open AI or Anthropic will end up knowing your own life better than Google, Apple, Microsoft, and even better than yourself at some point. For example, a friend of mine asked ChatGPT, tell me everything you know about me, and ChatGPT answered with a lot of facts about her life, her previous trips, her pets, and her partner. It’s quite scary and that’s just the beginning. You may wonder, is that a problem? In my opinion, it’s not a problem until it is one, which basically means, it’s a problem. All the data collected about everyone can either be leaked due to a hack, or worse, used under the pressure of governmental agencies.
Artificial Intelligence is transforming industries at a breakneck pace, and the entrepreneurs driving this innovation are at the forefront of this revolution. From groundbreaking applications to ethical considerations, these visionaries are shaping the future of AI. What does it take to innovate in such a rapidly evolving field, and how are these entrepreneurs using AI to solve real-world problems? As a part of this series, I had the pleasure of interviewing Sandoche Adittane.
Sandoche is a serial maker passionate about AI, Web3, open-source, design, and productivity. With an engineering background and a wide range of skills, he loves to learn, design, and build products from 0 to 1. He has released more than 45 products, including the brand new Llamao app (https://llamao.app), an offline and private AI ChatGPT alternative for mobile devices.
Thank you so much for joining us in this interview series. Before we dive in, our readers would love to learn a bit more about you. Can you tell us a bit about your childhood backstory and how you grew up?
I grew up in the countryside of France a few kilometers from Paris, in a small village with no more than one thousand inhabitants. My parents were pushing my brother, sister, and me to perform well at school so we were all good students. The fun part was that I spent most of my free time in front of my good old computer. At first I just randomly clicked on whatever was clickable (in an old Macintosh). Then, when I got a proper computer with Windows 95, I mostly used Paint and some games that were available here and there and in demos.
The real fun began when we started to get the 56k super slow internet. I started to use Microsoft Frontpage to build my first website. Thanks to some family friends I had the chance to learn to use a bunch of other software such as FL Studio, Adobe Premiere, After Effect, Photoshop, and a little bit of PHP and Mysql.
I was really enjoying crafting videos, making photomontages, and making more complex websites such as Breaking Out, a click-based online multiplayer game I built, inspired by the TV series Prison Break when I reached high school.
Can you share the most interesting story that happened to you since you began your career?
I think my first job had a quite interesting story to it. I first started an internship at IBM France as a UI/UX designer and software engineer (a mix of both). There, thanks to my mentor and manager I had the opportunity to build an internal website for IBM employees to share their skills and mentor each other. The platform was called CoachMe.
After my internship my manager offered me the opportunity to join the team as a full-time employee in the Netherlands and that’s how I got my first job. I was officially an IBMer, working in a small team of young people building internal tools.
When I started the job at IBM Netherlands I managed to continue developping CoachMe and implemented some of the features requested by my old internship manager who was acting as a product owner. But a few months later even though the project had quite a lot of regular users and a total user base of 30k employees, I was asked to sunset the project. The decision seemed to be based on political priorities rather than the project’s actual impact or popularity.
The good news is that I didn’t give up, I had to persevere and I managed to convince my new manager not only to continue the project, but also to grow the team bigger and to keep building for the users. In the end, we proudly got the Outstanding Technical Achievement Award, an internal award in the company. I believe that perseverance, believing in what you are building, and solving a real problem pays off and it’s worth fighting for that.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?
Good that we mentioned my internship and job at IBM, that’s where I met with Jérémie. He was at first my internship mentor and then later my colleague within the same team. Because I graduated as an Industrial Engineer I had a bit of imposter syndrome and lack of confidence working as a Software Engineer. But Jérémie really helped me grow, showing me what I was capable of, what I had to focus on next, and how to be a real software engineer with regular feedback and encouraging words. I am really grateful for that; it’s what put me on the path to becoming a full-stack engineer and led me to where I am now.
Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?
When I was a student at university, I was part of what we call in French, “Junior Entreprise”, which is basically a consulting club run by students offering services to companies, using other students as consultants and handling all the negotiations, paper-work, invoicing, payments… Thanks to this club and its partners we had the chance to attend some conferences and workshops and I remember in one of them one of the speakers said: “Done is better than perfect”.
It’s a simple sentence with a deep meaning. When you do something, basically anything, there is no need to overdo, no need to overthink it, no need to over-engineer. This “life lesson quote” had a big impact on me as a maker, especially when combined with another important one for me: KISS, Keep It Simple and Stupid. That’s how I managed to ship more than 45 products. When I build a product, I try to make sure that it solves ONE problem, so I can focus on ONE feature, and I will release it as soon as possible to test if it solves that problem and that users find that useful. If it does, and users pay for that, I’ll keep improving and supporting it. If not I’ll just drop it later.
You are a successful business leader. Which three character traits do you think were most instrumental to your success? Can you please share a story or example for each?
The first and most important one is ownership. When I work on a product, whether alone, with friends, or with a colleague I make sure to know who is responsible for what and most important of all: what I own. From there I can just trust myself, and eventually others to ship it. I take ownership and I am accountable for what I need to work on. This also translated in my past Engineering Manager role at Evmos, a web3 company. I made sure to assign owners for each part of the product and make sure they all knew about it.
Another very important character trait in my opinion is honesty, which usually goes along with transparency. Being transparent and honest with your peers makes communication healthy and creates a safe place for everyone to share how they feel. Also, being transparent and honest applies to clients and users. I always make sure to communicate bad and good news to users and be respectful and honest when they contact me for support. I noticed that talking honestly to your client removes unclarity and inaccurate expectations from them but also users will feel being listened to and understood. It’s something I do with all my responsibilities, all my projects but also with any other person. It just makes any relationship flow seamlessly.
Last but not least, antifragility is a trait necessary if you want to lead a business. It’s the ability to come back stronger after a stressful or hard situation. Without it I would never have been able to ship more than 45 products, I would have given up after the first failure.
Ok super. Let’s now shift to the main part of our discussion. Share the story of what inspired you to start working with AI. Was there a particular problem or opportunity that motivated you?
That’s a good question. The first time I saw that you could run Stable Diffusion locally on your own laptop I was amazed. It was the first time any person would be able to see photos, paintings, 3d renderings of anything or anyone without the internet. I was mind-blown. Who needs Google images?
Later it got super easy to run Large Language Models like ChatGPT locally on your device. Then I was wondering, if we can run it on a laptop can we run it on a phone? Then I did some research and I saw a few open-source tools to make this happen. Those tools or apps were quite painful to use for any regular user and using my designer mind I decided to build LLamao and make offline LLMs easy to use for anyone on their mobile device.
Describe a moment when AI achieved something you once thought impossible. What was the breakthrough, and how did it impact your approach going forward?
There were actually a few moments, first, what I previously mentioned using Stable Diffusion. Second, GitHub Copilot, integrated into Visual Studio Code (a code editor tool), was a major breakthrough. The code completion it offered was truly impressive and something I hadn’t imagined possible. However, at this point, it has largely been surpassed by Cursor, another innovative code editor. I’m confident that even more advanced tools will emerge in the near future.
The other mind-blowing AI moment was Dev-GPT from Jina AI, a tool where you can prompt anything and it will build an API for it and a tiny frontend plus some tests. For example, you can say “Generate a meme from an image and a caption” and it will create a UI where you can paste your image and write the caption, press a button and it will return your meme and even deploy it! That was super impressive back then. Recently, Anthropic released something similar called Claude code, even more flexible to build anything iteratively. I still need to take some time to play with it!
Talk about about a challenge you faced when working with AI. How did you overcome it, and what was the outcome?
I wanted to take advantage of AI tools to build Llamao, the offline AI App. To do so I wanted to use an AI programming assistant. I mostly used Cursor, particularly the inside feature called Cursor Composer. You type what you want, it tells what changes it will make with some explanation and you can approve or reject those changes.
While in theory that’s great, in practice it’s quite challenging to use because it’s not perfect. You still need to have good engineering skills to understand the problems and to explain them to the AI to fix them. At least for now!
Can you share an example of how your work with AI has had a meaningful impact (on others, on business results, etc)? What was the situation, and what difference did it make?
Our app Llamao lets users use AI directly from the phone without an internet connection and without storing any data. It targets two types of users.
The first type is the people looking for the offline aspect, for example, if you live in a remote place or a country with bad internet coverage. There are many countries where the Internet is not available everywhere and most people don’t own a computer. Llamao would bring them LLMs’ knowledge to help them in their daily life, and daily work, answer their doubts, and make them more productive.
The second type of user who benefits from our app is anyone with a question or something to share who doesn’t want their data collected. Llamao offers easy access to privacy: personal details, feelings, health concerns, financial data, every prompt is processed locally on the phone, and no data is stored anywhere!
Based on your experience and success, can you please share “Five Things You Need To Know To Help Shape The Future of AI”?
Coming from web3, an industry started by Cypher punks, I would say the most important fact that AI makers should keep in mind is: privacy matters. AI companies like Open AI or Anthropic will end up knowing your own life better than Google, Apple, Microsoft, and even better than yourself at some point. For example, a friend of mine asked ChatGPT, tell me everything you know about me, and ChatGPT answered with a lot of facts about her life, her previous trips, her pets, and her partner. It’s quite scary and that’s just the beginning. You may wonder, is that a problem? In my opinion, it’s not a problem until it is one, which basically means, it’s a problem. All the data collected about everyone can either be leaked due to a hack, or worse, used under the pressure of governmental agencies.
Second, I encourage makers, when possible, to leave part of what they are building open-source. It shapes innovation. That’s what DeepSeek did for example and we’ve seen recently Perplexity making an open source model that was actually based on DeepSeek. They took from the open source world and they gave back, that’s how innovation works at best!
Another important fact to keep in mind is, it’s super easy to run nowadays an AI on your device. A lot of these models are great and open-source. You can install Jan AI or LM Studio on your computer and run any ChatGPT for free, and on your phone, you have our app Llamao for example, or one of our competitors.
Moreover, thanks to the open source tool called exo you can even run heavier, and therefore smarter LLMs using the power of all your devices at home. This tool will share the load of performing the query to all the devices. So you could imagine a future where everyone would have their own super smart private assistant without risks of data leaks.
Lastly, while I’m not a fan of regulations, I believe that progress should ultimately benefit individuals. Having some guidelines or rules in place to prevent things from getting out of control would be beneficial. The paperclip problem is an ethical thought experiment that highlights the potential risks of AI. It imagines an AI whose sole goal is to maximize paperclip production. In its pursuit of this goal, the AI might ignore everything else, potentially leading to really bad consequences, including the destruction of human life.
When you think about the future of AI, what excites you the most, and how do you see your work contributing to that future?
I think what excites me the most is the rise of smart small LLMs as open source models such as Llama 3.2 or some of the DeepSeek variations. I am seeing that those small models are getting better and better and will probably at some point reach the level of centralized models such as gpt4o mini for example. My work will be to make those smart models available to the masses, giving power and privacy back to the people.
What advice would you give to other entrepreneurs who want to innovate in AI? Can you share a story from your experience that illustrates your advice?
Given that the AI space is going super fast, I would say: start now, do not wait a minute. Ship fast and iterate, use as much AI as possible to go faster. Also, keep in mind that whatever you build may be overtaken within a few months. Therefore it’s important to build something that can easily catch up with the competition or adapt to the evolution of AI technology.
When I started to work on Llamao, there were no competitors on Android and by the time I released there were already 3. Also, a few days after Llamao’s launch there were new very important models being released, such as DeepSeek. I had to add it to the app as soon as possible for the user to be happy!
Is there a person in the world, or in the US with whom you would like to have a private breakfast or lunch, and why? He or she might just see this, especially if we tag them. 🙂
I am a big user and a big fan of Perplexity, it makes my life so much more productive, therefore I would love to meet with Aravind Srinivas.
How can our readers further follow your work online?
My Twitter (@sandochee) or my personal website would be the best places to catch up on my work.
Thank you so much for joining us. This was very inspirational, and we wish you continued success in your important work.
Sandoche Adittane Of Llamao On Pushing the Boundaries of AI was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.