TrunkSubmittal Certification
for APM / PE.

One track from the AI-Enabled GC program, shown as the actual pieces an instructional designer produces — not a description of them. Built for a blended model: video, quizzes, and in-person components, with assessments that make certification mean something.

01

Track Overview

What this track is, who it's for, and how "trained, used in the workflow, and assessed" gets applied to this specific tool.

Learning objective

A reviewer can independently evaluate a TrunkSubmittal-flagged discrepancy against the cited spec section, confirm or overturn the flag, and either resolve it or escalate it correctly.

What the agent actually does

When a submittal is received, TrunkSubmittal analyzes it against the project's specifications and relevant RFI responses, flags non-compliant line items, and generates a structured recommendation. Everything in this track teaches someone to work with that output.

Certification standard

Trained

Modules 1–3 complete — understands what the tool checks and can classify flag types correctly

Used in workflow

Module 4 — reviewed real flags on their own project's live submittal log

Assessed

Module 5 — correctly resolves or escalates flagged discrepancies independently

Module sequence

01What TrunkSubmittal checks, and what it doesn'tVideo · 5 min
02Walkthrough of one flagged submittal — storyboarded belowVideo · 10 min
03Knowledge check — sample questions belowQuiz · 8–10 Q
04Live session on your own submittal logFacilitated · 30–45 min
05Certification assessment — rubric belowAssessed

Mechanism description and module structure grounded in a real customer deployment (public case study); ROI and adoption metrics from that source are intentionally left out — this track teaches the task, not the business case for the tool.

02

Module 2 Storyboard

"Walkthrough of one flagged submittal" — the actual pre-production document that would go to video production. No real TrunkSubmittal footage exists publicly, so each screen is a labeled placeholder for what would be a real product capture.

Frame 1 — Open
[PLACEHOLDER]
Trainer on camera, or title card
On-screen text "One submittal. One flag. Here's what you're actually looking at." Narration "You've got a submittal that came back partially compliant. Let's look at exactly what TrunkSubmittal caught — and what you still need to decide."
Frame 2 — Submittal received
[PLACEHOLDER — product screenshot]
Submittal log entry, status: "In Review"
On-screen text Submittal enters the log → routes to TrunkSubmittal automatically Narration "The moment this lands in your system, TrunkSubmittal starts comparing it against the spec section it's tied to — and any RFI responses that touch it."
Frame 3 — The flag
[PLACEHOLDER — product screenshot]
Flagged line item, cited spec section highlighted
On-screen text Flag: Product spec mismatch — Section 09 65 00 Narration "Here's the flag. Notice it doesn't just say 'non-compliant' — it shows you the exact spec section and what doesn't line up. That citation is what you're going to verify next."
Frame 4 — Your decision
[PLACEHOLDER — split screen]
Spec section (left) vs. submittal detail (right)
On-screen text Confirm the flag → draft the response. Overturn the flag → note why and proceed. Narration "This is the part TrunkSubmittal can't do for you. You're checking: does the flag actually hold up against the spec? If it does, you draft the response. If it doesn't, you note why and move on — but you don't just take the flag's word for it."
Frame 5 — Close
[PLACEHOLDER]
Trainer on camera, or recap card
On-screen text Next: try this on a real flag from your own project. Narration "That's the whole loop: flag, verify against the cited spec, decide. In the next session, you'll do this with a real submittal from your own log."
03

Sample Assessment

Two pieces of "assessments that make certification mean something": a knowledge-check question from Module 3, and the rubric an evaluator uses to certify Module 5.

Sample knowledge check — Module 03
Question 1 of 3 (sample)
TrunkSubmittal flags a waterproofing membrane submittal as partially compliant: the product model number matches spec, but the listed thickness is below the minimum called out in Section 07 13 00. What's the correct next step?
Approve the submittal, since the model number is the detail subcontractors get wrong most often.
Reject the submittal outright and ask the subcontractor to resubmit the full package.
Confirm the thickness gap against Section 07 13 00, then request updated product data for that item.Correct
Escalate to the architect directly, since any thickness deviation requires design review.
Sample knowledge check — Module 03
Question 2 of 3 (sample)
A flag cites an RFI response that appears to conflict with the spec section it's tied to. What should you do?
Trust the flag as written, since the system already checked both documents.
Follow the spec section only, since it's the controlling contract document.
Send the conflict back to the subcontractor and let them propose which document governs.
Read both documents yourself, since an RFI response can amend a spec and the two need reconciling.Correct

Module 5 — Certification rubric (live evaluator checklist)

CriterionPass
Correctly identifies whether each flag is compliant, partially compliant, or non-compliant
Cites the specific spec section (not just "the spec") when confirming or overturning a flag
Correctly overturns at least one deliberately-incorrect flag included in the assessment set, rather than approving it by default
Drafts a response to the trade partner that names the specific discrepancy, not a generic rejection
Completes all 5 flagged items within the session — no more than 1 requires evaluator prompting

The deliberately-incorrect flag (row 3) is the design decision that makes this assessment mean something — a reviewer who approves every flag without scrutiny fails, because the standard is judgment, not compliance with the tool's output.

04

Certification Tracker

How "trained, used in the workflow, and assessed" gets tracked in the LMS — sample learner records, not real data.

Sample data — illustrative only
LearnerRoleTrainedUsed in workflowAssessed
Learner 001APMCompleteCompleteCertified
Learner 002PECompleteCompleteScheduled
Learner 003APMCompleteIn progressNot started
Learner 004PEIn progressNot startedNot started

A real tracker would also roll these records up by account, cohort, and business unit — shown here at the individual-learner level since that's the record the certification standard is actually applied to.

ABOUT THIS SAMPLE

One track, shown as the real deliverables.

Built for the Instructional Designer & Learning Operations Lead role at Trunk Tools. Grounded in the company's actual product mechanism and a public customer deployment; placeholders are labeled everywhere real product footage or live data would be needed.