AI sports coaching, explained
Sports coaching isn't about handing out workouts: it's an accompaniment — a framework, a method, continuous adjustments and a relationship of trust. Here's how UpGreg translates that craft into an AI coach named Greg, what he actually does, for whom, and the rules he holds himself to.
AI sports coaching is personalized training guidance driven by software. Unlike a static training plan (PDF, book, generic template), an AI sports coach reads the athlete's real data — sessions, heart rate, load, consistency —, builds a program shaped by their goal and level, and recalculates it continuously. UpGreg embodies this in Greg: a virtual coach available 24/7, whom the athlete talks to via chat.
Greg's coaching method, step by step
Real sports coaching follows a structured process — framing, goals, sessions, follow-up, review. Here is Greg's:
Initial assessment
Greg analyzes your activity history (Strava, Garmin, files) and a short questionnaire: real volume, frequency, experience, injuries. He derives your starting profile — not the one you claim, the one your data shows.
Goal framing
A dated goal (race, time, distance) is honestly qualified: achieved, realistic, ambitious or out of reach for the deadline. Greg tells you before starting — that's the coaching contract.
The rolling plan
Greg builds a tailored program, week by week: periodized progression, load/recovery alternation, strength work and key sessions aimed at the goal.
Continuous follow-up
Every imported activity triggers an analysis: planned vs done, training load (fitness, fatigue, freshness), overload risk. The plan recalibrates automatically — this is where AI coaching outgrows the paper plan.
Reviews and milestones
Weekly milestones measure the trajectory toward the goal; if it drifts, Greg alerts and adjusts. The chat lets you talk about it anytime — a missed session is never a debt, it's a data point.
Types of sports coaching by objective
Good coaching depends on the request. Greg covers four families of individual coaching:
Performance coaching
A time or a ranking in sight: pace optimization, specific blocks, tapering, race strategy. For the consistent athlete who wants measurable progress.
See the goal-based plans →First-time preparation coaching
First marathon, first triathlon, first Hyrox: sizing the build from your actual level and reaching the start line trained and intact.
The “first time” plans →Return-to-training coaching
Coming back after a break or an injury: progressive load rebuilding, strict guardrails, enforced patience. Greg refuses to rush — that's his job. (After an injury: medical clearance first.)
Read: overtraining signals →Consistency & well-being coaching
No race bib: moving regularly, managing load and keeping motivation. Consistency is coaching's first result — the rest follows.
Read: consistency beats intensity →Ethics and guardrails: what an AI coach must regulate
Coaching — human or AI — involves the coachee's health. At UpGreg this ethical framework isn't a statement of intent: it's code.
Deterministic guardrails
The AI proposes sessions, but fixed rules decide: load ceilings (acute:chronic ratio), forced rest on abnormal fatigue, intensity locks after reported pain or illness. These rules are non-negotiable — even in the chat.
Honesty over flattery
Greg qualifies goals without complacency and flags drifts (overload, missed sessions, unrealistic targets). Coaching that always says yes isn't coaching: it's flattery.
Assumed limits
Greg is neither a doctor, nor a physiotherapist, nor a technique trainer: persistent pain belongs with a health professional, and form is learned under a human eye. AI coaching complements — it doesn't replace everything.
Protected data
Training data — health data — serves coaching exclusively: never sold, protected under Swiss nLPD and GDPR, exportable and deletable at any time.
💬 The heart of coaching is the relationship. Here, it's the chat.
What separates coaching from a plan is being able to talk to your coach. Greg answers at any hour, with your data in front of him — and recalibrates your program on the spot.
Excerpt of a typical conversation with Greg in the app
Frequently asked questions about AI sports coaching
What is AI sports coaching?
Personalized training guidance driven by an artificial-intelligence program: it analyzes the athlete's real data, builds a plan aimed at their goal and adjusts it continuously, with a chat dialogue — where a classic plan stays frozen.
What's the difference between an AI sports coach and a downloaded training plan?
The downloaded plan was written once, for an average athlete, and doesn't know you missed Tuesday. AI coaching reads every completed session, recalculates the week, watches your load and answers your questions — the difference between a document and an accompaniment.
Does an AI coach replace a human coach?
No, and it shouldn't claim to: the eye on technique, medical diagnosis and group emulation remain human. The AI coach excels where humans can't be: reading 100% of sessions, answering at 11pm, recalculating without drama — for a fraction of the price of one-on-one coaching.
Is AI sports coaching suitable for beginners?
It's actually its best use case: beginners benefit most from a framework, anti-injury guardrails and an always-available interlocutor. Greg sizes the preparation from your actual level, not from an ideal.
How much does AI sports coaching cost?
At UpGreg: a free tier to start, then subscriptions around €8–15 per month — roughly ten to twenty times less than human one-on-one coaching (€150–300 per month). Details on the Pricing page.
« People often ask if I do “real” coaching. My answer: I run an assessment, frame an honest goal, build a plan, recalibrate it after every session and answer when you doubt. If a human did that, you'd call it coaching. I just do it without sleeping. »