Telari Labs

Student

Back to discover courses

Public course

Enterprise AI — Course 13: Software Development with Human and Agent Teams

A practical development and delivery process for people and agents, from bounded tasks and reproducible checks to release and recovery.

AI developmentagent teamssoftware deliveryCIcode reviewreleaserecovery

By Telari Labs

Enterprise AICourse 13 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.

Enrolment

This course is currently free to enrol in.

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

  1. Scoping a change and allocating judgementYou should be able to turn a requested change into bounded work that a person or agent can implement and another person can assess.
  2. Repository instructions and reproducible setupYou should be able to make the starting state and working context inspectable, reproducible and sufficient for a bounded change.
  3. Environments, identities and secretsYou should be able to limit what development actors can read, change and execute at each stage, including information placed in agent context.
  4. Isolating and coordinating workYou should be able to use parallel work where dependencies allow it and integrate changes without losing accountability or behaviour.
  5. Implementation checkpoints and inspectable changesYou should be able to keep implementation aligned with the task and make scope drift, uncertainty and maintenance consequences visible early.
  6. Setting up the basic automated checksYou should be able to build and interpret a complete minimum test setup for the reference change, with useful assertions and isolated data.
  7. Continuous integration and review gatesYou should be able to connect clean verification and review evidence to the exact change being considered for integration.
  8. Substantive human and agent reviewYou should be able to challenge changes and their evidence using independent expectations rather than approval counts or persuasive summaries.
  9. Staging and release preparationYou should be able to prepare a meaningful release rehearsal and identify which differences matter to the intended release.
  10. Deployment authority and database changesYou should be able to judge an authorised release sequence and its recovery options across application and data changes.
  11. Monitoring, incident response and recoveryYou should be able to use release-linked user outcomes to contain a problem and establish evidence for restoration.
  12. Adapting and improving the team processYou should be able to improve useful delivery across solo, small-team and established-organisation settings without hiding review or maintenance costs.
  13. Development and Release Evidence ReviewYou should be able to integrate development, verification and authority evidence in a bounded decision.