Enrolment
This course is currently free to enrol in.
Public course
Enterprise AI — Course 11: Designing Bounded Agents
Design choices for agents and simpler workflows, with explicit goals, tools, memory, supervision and operating limits.
agentsworkflowsplanningtoolsauthoritymemorysupervisioncoordination
By Telari Labs
Enterprise AICourse 11 of 25
Learning on Telari
Created by Telari Labs, delivered through Telari
The course creator controls the material and assessment standards. Telari provides structured lessons, practice, feedback against those standards and a record of your progress. See how Telari works.
AI access
The first managed lesson is included on Free. Continue with Plus or your own supported API key; premium assessments and certificates may require Plus.
Audio
No reviewed lesson audio is currently advertised for this course.
Lesson list
- When agentic choice is justifiedArchitecture choices grounded in control flow, task variability, total effort and decision-relevant uncertainty.
- Goals, plans and stopping rulesGoals and plans with observable completion evidence, explicit scope, resource limits and bounded incomplete outcomes.
- Tools, observations and authorityTool selection and observation grounded in contracts, source trust and enforceable action authority.
- Memory and continuityRetained information supports continuity when its purpose, provenance, scope and currency remain visible.
- Supervision that can workReview evidence, approval gates, capacity and intervention controls make supervision operationally useful.
- Single and multiple agentsAgent decomposition follows dependencies, evidence origins, accountable coordination and total effort.
- Defining the operating envelopeA coherent operating envelope connects useful choice, authority, evidence, limits and supervision, and changes when a material condition changes.
- Final assessmentNew cases test whether bounded design decisions change appropriately with new evidence, constraints and authority.