Train To Mountain Launches the First Training Algorithm Built Around the Demands of a Specific Summit or Trek

Train To Mountain announced the launch of its personalised training platform for mountaineers and trekkers, built on a proprietary algorithm that plans training backwards from the demands of a specific summit or trek.

Trekking plans go live in November, extending the engine from classic summits to multi-day treks worldwide. Mountaineering and trekking place more on the body than almost any other amateur sport: hours of sustained climbing, long descents on tired legs, altitude, weather, back-to-back days and the mental resilience to keep moving through all of it.

Yet while runners, cyclists and Hyrox athletes train with data-driven apps that adapt to them, mountain athletes still borrow plans from other sports, buy static PDF programmes, or piece together advice from experts who disagree with each other online. “Imagine someone training for a marathon with a PDF plan today. It is hard to picture, and yet that is still the norm for people preparing for a 4,000 metre (13,100 ft) summit,” said Jakob Ulcnik, founder of Train To Mountain. “One-to-one coaching works, but it costs more than most people can justify, and it was built for the expedition climber, not the weekend mountaineer with a job and one mountain they cannot stop thinking about.”

The gap Train To Mountain fills sits between those two options. The team spent three months building an algorithm specifically for mountaineering and trekking, rather than adapting one from endurance sports. It starts from the mountain: every one of the 60+ summits and treks in the catalogue, from Mont Blanc to Aconcagua, carries a demand profile covering summit-day hours, vertical gain, descent, altitude, pack weight and terrain. The athlete’s plan is built backwards from that profile and their date, using the training history from their own wearable.

The engine tracks six dimensions of readiness that general fitness apps do not measure: ascent, descent, long days, back-to-back days, consistency, and aerobic fitness.

A fitness-fatigue model from endurance science plans the load across multi-week timelines. Heart-rate variability from the athlete’s watch gates how hard each day gets. An audit layer rejects any plan the body or the calendar cannot absorb, including weekly load caps and ascent-rate feasibility checks.

The plan recalibrates every week from what the athlete actually did. “Anyone can wrap a chatbot around a workout. The hard part is the mountain,” Ulcnik said. “We hold verified demand data for every peak we serve, so the plan is personalised twice: once to the person, through their wearable data, and once to the mountain they are going to stand on. Language models write the words. The maths decides the load.”

The platform adapts to beginners and experienced mountaineers alike, in both urban and hilly environments, and is designed around real lives: sessions can be rescheduled when the week changes, buddy training lets two or more people prepare for the same objective together, and the Summit Readiness Score shows the athlete which of the six dimensions is holding them back and what raises it. Anyone can take a free readiness check for their peak before signing up. Train To Mountain connects directly to Garmin, Suunto, Coros, Polar, and Apple Health.

Guiding companies share the platform with their clients through partner links, so people arrive at the mountain prepared. Climbing coach Eric Hörst featured the platform on Training For Climbing in September 2026. Ulcnik, a Slovenian alpinist based in Amsterdam with more than ten years in technology, started building the product after his own experience in December 2025, when months of training with PDF plans and online advice left him unprepared on the way to a ridge at about 3,500 metres (11,500 ft). The platform went live in September 2026 with technical co-founder Gino Kok.