Learn it

Conversational Instruction

Your subject-matter expert, available for every learner, every time.

Runs a live instructional dialogue with one learner — diagnosing what they don't yet grasp, explaining at the right level, and adjusting each turn based on their response. Reach for this when you want the quality of a good tutor at the scale of a platform.

Shape

LIVE 1:1 SESSIONLearnerTutor agentdiagnose · explain · check“why does this matter?”explains, pitched to levelattempts an answerchecks, rescaffolds next turngroundingtranscript

Operational dimensions

Human co-pilot

Person and system work side-by-side.

On demand

Fires when a user asks.

Low data gravity

Light state; replaceable any time.

Closed surface

No external systems on either side.

Inputs

  • learner utterance + session history
  • grounding material (curriculum doc, syllabus, role context)
  • optional prior performance signal

Outputs

  • instructional response pitched to learner's current level
  • in-session understanding signal (per-turn)
  • session transcript for downstream coaching or assessment

Mechanism

Conducts turn-by-turn 1:1 dialogue with a learner — diagnoses gap, explains, checks understanding, scaffolds the next turn — all WITHIN a single session.

Why this is a primitive

Cannot be decomposed — the diagnose-explain-check-rescaffold loop is a single conversational operation that happens at the granularity of one utterance. Routing across sessions, scoring of a fixed item, and assessment-of-mastery are different primitives that this one does not perform.

Where it shows up

Professional training firm onboarding new hires — AI tutor walks each hire through product knowledge, adjusting depth to their background
K-12 edtech platform — student asks why photosynthesis matters; system diagnoses misconception, explains with an analogy, checks comprehension
Enterprise compliance team — employee asks about a specific policy edge case; system explains the rule and tests understanding before moving on

Related primitives

Tags

AIlearningconversationalhuman-collaborativereal-time

See where it fits.

Primitives are configured into named solution shapes for each client’s domain. The fastest next step is a conversation about which shape fits your problem.

Start a conversation