A
- Accelerated Expertise Training
- Skillgym's learning method: practicing expert strategies in realistic conversations, with guidance and feedback, to shorten the path from knowing what to do to doing it under pressure.
- AI Live Coach
- The in-session coach that suggests the next move while the conversation is happening, calibrated on the trainee's level: fully guided for beginners, strategic for experienced people.
- AI Role Play
- A conversation with an AI character who has a personality, needs and a standpoint, and reacts to what the trainee says. In Skillgym it is governed by a methodology, not left to the model.
- Authoring Tool
- The guided pipeline that turns a client's materials, playbooks and competency model into a circuit of scenarios in under an hour.
B
- Behavioral metrics
- Measures of the behaviors expressed in what the trainee said, graded in context and aggregated on the client's competency model. More than 50 behaviors are measured.
- Bootcamp
- A training program with a defined population, duration (typically 4 to 8 weeks), set of scenarios and number of sessions per person, run on the training protocol.
C
- Character (digital human)
- The AI interlocutor in a scenario: a fictional person with a background, explicit and implicit needs and a personality built on recognized psychometric models, rendered as a real-time digital human.
- Character feedback (emotional feedback)
- The character's own account, in first person, of how they lived the conversation. It describes the character's simulated reactions as determined by the scenario; it is never an analysis of the trainee's emotions.
- Circuit
- A set of synergic scenarios, typically four, designed around the same competencies in different situations, to build cognitive flexibility rather than a single response.
- Cognitive flexibility
- The ability to apply the same skills in new, unexpected situations. Built in breadth, across varied scenarios.
- Competency mapping
- The link between the behaviors measured in conversations and the client's own competency framework, with the client's structure and weights.
G
- Governed AI
- Generative AI operated inside a layer of proprietary algorithms that keep the character consistent, recognize the strategy expressed in the trainee's sentences, align coaching with the expert model and verify outputs before they reach the trainee.
H
- Human-in-Command
- The principle that people design the method, validate the content and use the results; the AI makes the conversation real and decides nothing about people.
L
- Learning agility
- How fast and how flexibly a person adapts skills to new contexts and retains them over time.
- Live Avatar
- A ready-made digital human from the Skillgym library, usable across scenarios; custom avatars can be created for a client.
- Localization
- The adaptation of a scenario to a target language and culture before the session, so the conversation runs natively; available in more than 30 languages.
M
- Multiple valid paths
- The design principle that several different moves can resolve the same strategic challenge in a conversation. The coach guides towards the strategy; the trainee finds their own words.
N
- Non-self-learning
- The property that the system does not change its criteria from the sessions it runs. Changes are human decisions, versioned and released.
P
- Practice Fastlane
- Skillgym's framework: capturing the expert model, guiding practice, guiding the debrief and orchestrating the circuit, so that people reach the expert's approach in a fraction of the time of unguided trial and error.
- Preparation room
- The pre-session phase in which the trainee reads the context, the character and the objectives, organized the way an expert would read the situation.
- Program dashboard
- The view available to program owners and tutors: participation, reliability, aggregate improvement trends by team and over time. Not a view of individual debriefs.
Q
- Quality of the strategy
- Skillgym's main result metric: the overall quality with which the skills required by a conversation are used, in context. It is the metric that builds a person's confidence in handling critical conversations.
R
- Reliability
- A discipline metric: the share of scheduled sessions a trainee actually attends. Followed at program level.
S
- Scenario
- A conversation to practice: situation, character, objectives, strategy, hints and feedback criteria. Ready-made from the catalogue or built from the client's materials.
- Self-awareness
- The distance between how a person assesses their own performance and how it was actually evaluated, internal (own impact) and external (how the other person perceived it). 84% of professionals enter with low or intermediate self-awareness.
- Session
- One practice conversation with its preparation and debrief, around 20 minutes.
- Smart Debrief
- The post-session analysis: 20+ behavioral metrics, turn-by-turn commentary, knowledge and compliance pointers, and an interactive coaching chat with rehearsal of the critical moment.
T
- Training protocol
- The rules that set how often, how long and on what each person practices, and adapt the schedule to attendance and results.
- Transfer
- The improvement of real performance as a result of practice. Validated independently in a controlled study at USI Lugano.
V
- Velocity
- The automaticity with which effective responses are produced. Built in depth, through repetition on a few scenarios with growing difficulty.