EGYM: The unstaffed gym problem
カートのアイテムが多すぎます
カートに追加できませんでした。
ウィッシュリストに追加できませんでした。
ほしい物リストの削除に失敗しました。
ポッドキャストのフォローに失敗しました
ポッドキャストのフォロー解除に失敗しました
-
ナレーター:
-
著者:
Recorded one day before EGYM announced its merger. Co-founder and CTO Florian Sauter on why the AI is not a language model.
Recorded in January 2026.
Every fitness technology conversation in 2026 starts with large language models. Florian Sauter runs the technology at EGYM and says that is not what the company builds with. On his account the core product sits on classical machine learning over structured data, and the hard input is not workout data at all. It is hard assessment data from equipment.
Sauter is co-founder and chief technology officer at EGYM, the Munich connected-equipment and gym software business. He has held the role for around fifteen years, from a single room to a company that on 31 March 2026 completed a merger with Playlist, the parent of Mindbody, ClassPass and Booker, in a deal the companies put at 7.5 billion dollars enterprise value with 785 million dollars of new equity. This conversation was recorded on 14 January 2026, the day before that merger was announced.
In this conversation:
- Why EGYM Genius runs on classical machine learning rather than large language models
- Why assessment data, not workout data, is the input that closes the loop
- What member care looks like in a gym segment that carries no staff
- Why a training plan has no single outcome to optimise for
- What combining hardware and software actually buys a product team
Links:
- Florian Sauter on LinkedIn
- EGYM
- Natalia Karbasova on LinkedIn
- Tomorrow Health
- Apply to join FitTech Club
Chapters:
00:00 The gyms with nobody on the floor
00:25 Three things to listen for, and what happened the next day
01:37 Fifteen years, from a single room
02:49 EGYM Genius and the average member
03:55 Why assessment data is the hard part
05:04 Less generative AI, more machine learning
06:54 The Fitness Hub and computer vision
07:33 A training plan has no single right answer
08:53 Everything in house
09:11 Member care with no staff on the floor
10:24 Copilots, and mixed results
11:15 What hardware plus software actually buys you
12:09 Fitness converging with health
13:41 The problem AI will never solve
14:06 Build value, technology follows