Two Systems, One Platform
keysessions.ai is built on a deliberate separation. The periodization engine handles the structural decisions — load targets, session distribution, progression curves, readiness calculations, and adjustment logic, producing over 45 million valid output combinations. The AI operates alongside it as an interpretation layer — a coach that reads the data, understands the context, and communicates what matters to you. The engine computes. The AI explains. Neither crosses into the other's domain.
What the Engine Does
Every structural decision in your training comes from the periodization engine. It decides how load is distributed across the week. It determines which sessions carry high intensity and which are recovery. It calculates your readiness after every completed session and decides whether the plan holds, adjusts, or protects you. It enforces progression rules — conservative or aggressive, depending on your profile. These are not suggestions from a language model. They are computed outputs from an algorithm built on training science, tested against real training scenarios, and designed to be transparent in how they work.
What the AI Does
The AI gives your data a voice. When you open the dashboard, a daily briefing tells you where you stand — what day of the block you're in, how your fatigue is trending, what today's session is designed to achieve, and what to keep in mind. After a session, the AI analyzes your execution in the context of the plan and delivers insight that goes beyond numbers. When you have a question — about your progression, about a specific session, about how a missed day affects the block — the coach answers with your data, not with generic training advice pulled from a textbook.
Why the Separation Matters
AI is exceptional at interpreting context, generating natural language, and finding patterns in data. It is not reliable for enforcing safety constraints, maintaining load invariants, or making decisions that must be consistent and reproducible. A language model that decides your TSS target might give you a reasonable number today and an unsafe one tomorrow — with no way to predict which. KSN doesn't take that risk. The engine owns the math. The AI owns the narrative. The athlete gets both: computational precision in the plan, and intelligent interpretation in the coaching.
The Coach Experience
The AI coach is not a chatbot bolted onto a training log. It's an integrated layer with full visibility of your training state — your profile, your mesocycle, your session history, your readiness, your progression. Every response is grounded in your actual data. For example, it knows that you're on day 5 of the second week of a build mesocycle, that your last two sessions were rated harder than expected, and that tomorrow is a scheduled recovery day. That context shapes everything it tells you. The experience is designed to feel like talking to a coach who has read your entire training log before the conversation starts.