When you select your training focus and programming tendencies in Personal trAIner PRO, you're not filing paperwork. Those settings load specific instruction modules into the session-building process — modules that shape the exercise selection logic, the session structure, and the way the AI frames each workout. Understanding what they actually do makes it easier to set them correctly and to diagnose why output looks the way it does.
Training focus: what it tells the system about your practice
Training focus captures the type of work you primarily do. The options are strength, hypertrophy, athletic performance, weight loss, general fitness, older adults and functional training, and rehab and corrective. You can select up to three.
Your training focus tags are fed to the session-building process as context data. They tell the system what kind of trainer you are and what kind of athletes you typically work with. This shapes the exercise categories the system weighs most heavily when it's building a shortlist of exercises for a given session — a trainer tagged for athletic performance will see more power-based and multi-directional work weighted toward the top of the shortlist than a trainer tagged for older adults and functional training, even for an otherwise similar client profile.
One thing worth understanding: training focus tags don't override client-level goals. If a client profile specifies weight loss, that goal is factored in regardless of whether weight loss is in your training focus. The focus tags describe your practice; the client profile describes the individual. The system reads both.
If you work with a genuinely mixed roster, select the two or three tags that represent the majority of your work. Selecting all seven provides less useful signal than selecting the two or three that actually characterize what you do.
Programming tendencies: where the real session structure comes from
Programming tendencies are different from training focus in a meaningful way. Each tendency you select loads a specific instruction module — a set of detailed generation rules — into the session-building process. These aren't labels the AI uses loosely for flavor. They're operational instructions that directly change what gets built.
The available tendencies are straight sets, supersets frequently, two main lifts per session, and conditioning built into the session.
Straight sets
Selecting straight sets loads an instruction module that tells the system to structure work as discrete sets of a single exercise, with rest between sets, before moving on. This is the most common structure in traditional strength and hypertrophy programming — you finish all sets of exercise one, rest, then move to exercise two. The module also adjusts how the system thinks about session pacing and exercise sequencing.
Supersets frequently
The supersets module tells the system to pair exercises and generate them in alternating format. When this is active, the Step 2 shortlist is built with pairing logic in mind — exercises are selected and grouped so that the superset format makes sense structurally, rather than just tagging random exercises as paired after the fact.
If you use supersets for specific purposes — antagonist pairing, time efficiency in shorter sessions — you can complement this with your philosophy statement to add more nuance than the toggle alone provides.
One thing to note: straight sets and supersets frequently describe different session structures. Selecting both sends contradictory instructions into the same generation process. If your actual practice is "mostly straight sets but I'll superset accessories," that nuance belongs in your philosophy field rather than in both tendencies selected simultaneously.
Two main lifts per session
This tendency changes the shortlist-building logic in Step 2 specifically. When it's active, the system builds the exercise shortlist with two compound movements designated as primary lifts, rather than one main lift followed by supporting work. This affects which exercises are prioritized for the primary slots and how the rest of the session is structured around them.
For trainers whose programming is built around a primary and a secondary compound — say, a squat pattern followed by a hinge pattern, or a press followed by a pull — this tendency ensures that structure is reflected in what gets generated without you having to specify it for every client.
Conditioning built into the session
When this is active, the system adds a finisher to the session structure. The finisher is generated at the end of the session and can take several forms depending on the client's goals and the session context — a short metabolic circuit, a single high-effort movement, a simple conditioning piece. The conditioning tendency doesn't determine the finisher's exact content; it tells the system that a finisher slot should exist and should be generated as part of the session.
Setting these correctly for your actual practice
The most useful way to think about your programming tendencies is to describe a typical session you'd write by hand and then select the tendencies that match it. If your typical session is a main compound, three or four accessories in straight sets, and nothing else — that's straight sets with one main lift. If your typical session has a main lift, a secondary compound, and then paired accessories — that's two main lifts and possibly supersets frequently. If you always end with a conditioning piece — add conditioning built into the session.
If you find that generated sessions are consistently structured differently from how you'd write them yourself, the programming tendencies are the first place to check. They're the settings most directly connected to session format.