Building a submittal log for a Division 26 job is one of the most time-consuming tasks in an electrical subcontractor's preconstruction workflow. It's also one of the most automatable.
Here's a breakdown of how AI submittal software approaches electrical submittals — what it automates, what it doesn't, and what you need to implement it on your jobs.
the electrical submittal problem
Division 26 specs are dense. A commercial or institutional electrical job might have 80–120 submittal line items — switchgear, panelboards, lighting fixtures, conduit fittings, wiring devices, fire alarm components, low-voltage systems, motor controls.
For each of those line items, your PM needs to:
- Find the relevant spec section
- Identify the specified product and any approved-or-equal options
- Check what the engineer has approved on past jobs
- Select the right product from your suppliers
- Document it correctly in the submittal log
That process takes 10–20 minutes per line item when done manually. On a 100-item electrical submittal, you're looking at 15–30 hours of PM time before the first submittal goes out the door.
Multiply that across 10–15 jobs per year and you have an entire person's worth of time spent on submittal prep.
what ai submittal software does for division 26
AI submittal software automates the research and matching phases of this process. Here's how it handles each step:
Reading the spec book The AI ingests your project's Division 26 spec book and extracts the product requirements for every submittal line item. It identifies specified products, approved manufacturers, substitution language, and any engineer notes embedded in the spec.
This alone saves significant time — most PMs flip back and forth through a spec book manually, re-reading sections to make sure they haven't missed anything. The AI does a single complete pass.
Matching against approval history The AI cross-references the spec requirements against your company's submittal history — every product that's been approved on past jobs, organized by engineer, firm, and job type.
This is the high-value step. If you've done five hospital jobs with the same consulting engineer and they always approve a specific manufacturer for panelboards, the AI knows that. It doesn't rely on your PM remembering. It pulls from the database every time.
Generating the draft log The output is a submittal log with product selections filled in — ready for your PM to review rather than build from scratch. The PM's job becomes quality control, not data entry.
what still requires human judgment
AI submittal software doesn't eliminate your PM's role — it changes it.
Things that still require human review:
- Substitution requests: When the AI recommends a product that isn't the specified manufacturer, your PM needs to decide whether to request a substitution and how to frame it.
- Value engineering opportunities: The AI optimizes for approval likelihood, not necessarily for cost. Your PM might want to override a recommendation to go with a less expensive option that's still spec-compliant.
- New engineers: If you're working with an engineer your company hasn't worked with before, the AI has limited history to draw from. Your PM's knowledge fills that gap.
- Ambiguous specs: Some Division 26 specs are written loosely. The AI will flag these, but the judgment call on product selection is still yours.
The AI is excellent at the 80% of line items that are straightforward. The remaining 20% — the edge cases, the substitutions, the ambiguous specs — still need your PM. But your PM gets to spend their time on those 20% instead of all 100%.
what you need to get started
Getting AI submittal software running on your electrical jobs requires three things:
1. Historical submittal data The quality of the AI's product recommendations is directly proportional to the quality of your approval history. If you have past submittal logs — even in PDF or Excel format — those can be imported to seed the system.
Most electrical subs have this data somewhere. It's not always organized, but it exists. The implementation process typically involves ingesting and normalizing that historical data.
2. A spec book upload process You need a way to get Division 26 spec books into the system. This is usually a simple file upload — the AI handles the parsing. If your GC provides specs digitally (which most do for commercial jobs now), this is straightforward.
3. An export format that fits your workflow The AI-generated submittal log needs to drop into the format you actually use — whether that's a specific template your GC requires, your company's standard Excel format, or a project management platform like Procore.
Good AI submittal software is built to export into existing formats rather than forcing you to change your workflow.
the roi calculation for electrical subs
Here's a simple way to think about the ROI:
- Average hours per job on submittal prep (manual): 20 hours
- Average PM cost: $45–65/hour fully loaded
- Cost per job (manual): $900–$1,300
- Jobs per year: 12–15
- Annual submittal prep cost: $10,800–$19,500
AI submittal software typically reduces that time by 70–80%, meaning you're looking at $7,500–$15,000 in annual time savings for a mid-size electrical sub running 12–15 jobs per year. The ROI calculation is usually fast.
The less-quantifiable benefit — fewer rejections, less resubmittal delay, lower engineer friction — adds further value on top.
how kjags advisors implements this for electrical subs
Our AI submittal employee for electrical contractors is purpose-built for Division 26. It was developed specifically for the Baltimore and DC market, where the combination of institutional work (hospitals, universities, government) and high submittal volume makes the time savings most significant.
The implementation process:
- We ingest your historical submittal data
- We configure the system for your standard product preferences and suppliers
- We run a pilot on an active job spec — usually one of your current projects
- You review the output against what you would have produced manually
- We adjust based on your feedback and deploy
Most electrical subs are using the system on live jobs within a week of starting the implementation.
If you want to see what the AI submittal employee produces on an actual Division 26 spec, book a call with us and we'll run a live demo on one of your current job specs. No slides. Just the tool working on your actual data.
kjags advisors builds AI employees for construction companies in Baltimore and Washington DC. Our AI submittal employee for electrical contractors handles Division 26 — spec reading, history-based matching, and submittal log generation.