Course 4.9 · Curriculum 2026.1
The autonomous task loop: the org is the queue and the completion target
Point an agent at the work already sitting in your org. Read the open task records, do the work, verify it under the real user, delete the test data, and only then write back what was done and how it was checked.
- Estimated effort
- 4 hr
- Lessons
- 34
- Modules
- 6
- Material
- 4 hr
Estimated effort is the sum of the lesson and lab times in the course outline, for this version of the course.
This course is planned and not on sale. The curriculum below is the plan, and it may change before the course opens.
See the courses on sale01What you will learn
7 capabilities, stated as things you can do
Each one is what you should be able to do at the end, not what the course covers. If an outcome is not testable, it is not an outcome.
- Design a loop where the CRM task is both the queue and the completion record.
- Split tracking three ways: the canonical queue, the working board, and the reply the requester actually receives.
- Write a completion note that says what was done and how it was verified, in the requester’s language.
- Stop the loop safely, so a half-finished change never sits in an ambiguous state.
- Require the agent to propose improvements rather than only clearing the list it was given.
- Keep an audit trail that survives someone asking who changed this in March.
- Decide which task types an agent may never pick up on its own.
02Curriculum
6 modules, 34 lessons
Every lesson shows its length and its type, from the course outline.
6 modules · 34 lessons · 4 hr of material
An agent starts each run against Salesforce itself: the open work is already sitting there as task records.
- VideoSalesforce as the task queue and the completion targetPreview8 min
- VideoReading open work from task records, not from a chat history7 min
- VideoWhy the CRM task is the canonical record, not the sidebar6 min
- ReadingWhat “open” means precisely, before an agent trusts it5 min
- LabLab: read the open-task queue for a real object10 min
- QuizCheck: the org as the queue8 min
The CRM task is the canonical queue and completion record. A project tool is the working board. A customer-facing mirror carries the reply.
- VideoThe CRM task, the working board, and the customer-facing mirror7 min
- VideoThe CRM task as the canonical queue and completion record6 min
- VideoA project tool as the working board for sub-issues6 min
- ReadingA plain-language reply the requester can be sent verbatim5 min
- LabLab: mirror one task across all three tiers10 min
- QuizCheck: three-tier tracking8 min
Doing the work is the easy half. Nothing is marked complete until it is verified under the real user’s permissions and the test data is gone.
- VideoDoing the work, then verifying it under the real user’s permissions7 min
- VideoCleaning up the test data before anything is marked complete6 min
- ReadingThe rule from course 4.2, applied on a loop5 min
- LabLab: run one task through pickup, work, verify and cleanup10 min
- QuizCheck: doing the work and verifying it8 min
- VideoWhat happens when verification fails partway through6 min
A completion note has to say what was done and how it was checked, written so the requester can read it without translation.
- VideoA completion note that says what was done and how it was verified7 min
- VideoWriting for the requester, not for the next engineer6 min
- ReadingThe note that claims success and proves nothing5 min
- LabLab: write a completion note for a task you closed9 min
- QuizCheck: writing the completion note8 min
- VideoWhy the note is the record, once the task closes6 min
A loop that stops badly leaves a change half-finished and ambiguous. This module is the stop conditions and the do-not-touch list.
- VideoStopping the loop safely, so nothing sits half-finished7 min
- VideoDeciding which task types an agent may never pick up6 min
- ReadingContinuously proposing how to make the org better, not just clearing the listPart of the job is proposing improvements, not only clearing the given queue.5 min
- LabLab: define the stop conditions for an unattended run9 min
- QuizCheck: stopping safely8 min
The intake rule from the same system: one message per issue for its whole life, and personal news as a direct message, never a channel line.
- VideoOne message per issue, for its whole life, edited in placeAn issue gets exactly one message in its channel for its whole life. The card carries its own state.7 min
- VideoWhy the card carries its own state instead of adding new lines6 min
- ReadingNews about one person arrives as a direct message, not a channel line5 min
- LabLab: audit a task’s history back to who changed it in March10 min
- QuizCheck: the autonomous task loop8 min
04Before you start
What you need first
Assumed knowledge and setup
- Courses 4.1 and 4.2. The done rule is what keeps this loop safe.
- An org with a task or issue object you may write to, plus a sandbox to rehearse in.
- A coding agent you can run unattended, and somewhere to log what it did.
05Who teaches it
Elliot Saha
Co-founder, Chief Technology Officer
Platform architecture, custom Apex and Lightning Web Components, AI systems, and the Academy platform itself.
06Reviews
No learner reviews yet
This course has not been taken by enough people to publish an honest rating, and we will not print invented quotes on a page that sells verification discipline. Here is what we can evidence instead.
- Assessment
- Lesson checks onlyThis course carries lesson checks and issues no certificate.
- Curriculum version
- 2026.1You enrol into a version. Content changes do not move you mid-course, and your required-lesson count is snapshotted at enrollment.
- Refunds
- 14 daysAn individual purchase can be refunded within 14 days if you have completed less than 20% of the lessons. Team seats are refundable before they are assigned. From the Academy terms.
Start 4.9
This course is in development and not yet open for enrolment. The courses open now are listed in the catalogue.
- 4 hr estimated effort, labs included
- A ShareCo certificate is not a Salesforce certification