Minimalist feature image showing AI workflows in Excel with clean blue UI elements and structured process visuals
By Jake Heller March 19, 2026 AI & Technology

How to Build Repeatable AI Workflows in Excel with Shortcut Skills

The biggest frustration I hear from CRE professionals using AI in Excel is inconsistency. You can run the same deal twice and get completely different outputs. However, an AI underwriting skills workflow solves this by standardizing how every deal is analyzed and formatted.

Instead of unpredictable results, you get consistent models every time.

The Problem: Inconsistent AI Outputs

AI works well for quick analysis. However, inconsistency becomes a real issue when:

  • You run multiple deals weekly

  • Your team needs standardized models

  • Investors expect consistent reporting

  • Junior analysts rely on structured outputs

As a result, inconsistent formatting slows down workflows and creates confusion.

Infographic showing inconsistent AI outputs, causes like no standard models, and workflow impact such as delays and confusion
Inconsistent AI outputs create delays, confusion, and extra work without standardized workflows in place

What Are Shortcut Skills?

Shortcut Skills are repeatable workflows built inside Shortcut AI.

How They Work

  • You define a trigger

  • You set the behavior

  • You attach your template

  • The system remembers everything

Once set up, the skill runs automatically when triggered. Therefore, you no longer need to rewrite prompts or explain your format each time.

Why Skills Are Different from Prompts

A normal prompt works once. A skill works every time.

  • Prompts require repetition

  • Skills store preferences

  • Prompts vary in output

  • Skills ensure consistency

As a result, skills are better suited for repeatable workflows like underwriting.

How I Built a Multifamily Underwriting Skill

The setup process was simple and took about 10 minutes.

Step 1: Define the Trigger

I set the skill to activate whenever I am asked to underwrite a multifamily deal.

Step 2: Define the Behavior

I instructed it to:

  • Extract data from an OM

  • Build the underwriting model

  • Match my exact template structure

Step 3: Attach the Template

I uploaded my Excel underwriting model. The system analyzed:

  • Deal scorecard

  • Property assumptions

  • Acquisition assumptions

  • Return metrics

  • Analyst notes

Step 4: Answer Follow-Up Questions

The system asked key questions before finalizing:

  • Should formulas be dynamic? → Yes

  • Should it solve for the max purchase price? → Yes

  • Should analyst notes always appear? → Yes

  • Use standard renovation estimates? → Yes

These questions improved the output significantly.

Minimal infographic showing four steps to build an AI underwriting workflow in Excel with simple icons
A simple four-step process to create a consistent AI underwriting workflow using triggers, behavior, templates, and rules

Testing the AI Underwriting Skills Workflow

I tested the skill on a real multifamily deal in Los Angeles.

What Happened

  • Uploaded the OM

  • Skill extracted all data

  • Built a full model

  • Solved for max purchase price

The output closely matched my template.

What Worked Well

  • Correct structure across all sections

  • Proper placement of assumptions

  • Clean deal scorecard

  • Consistent formatting

What Needed Improvement

  • Analyst notes column was missing

  • Some formulas were not fully dynamic

Both issues can be fixed through iteration.

Minimal infographic comparing AI underwriting workflow results showing what worked and areas needing improvement
A quick test shows AI underwriting can deliver consistent results with minor areas to refine

Standardization vs Speed

Speed is helpful. However, standardization is more important.

Without Skills

  • Different formats every time

  • Inconsistent layouts

  • Variable output quality

  • Harder team collaboration

With Skills

  • Same format every time

  • Consistent structure

  • Predictable outputs

  • Easier scaling

Standardization Impact Table

Factor Without Skills With Skills
Output Format Inconsistent Standardized
Analyst Efficiency Lower Higher
Investor Reporting Variable Consistent
Workflow Speed Slower Faster

Comparing Shortcut Skills to Other Tools

Claude for Excel

  • Strong analysis capabilities

  • No persistent workflows

  • Requires repeated instructions

Index AI

  • Workflow features available

  • Less intuitive setup

Shortcut Skills

  • Persistent workflows

  • Template-aware outputs

  • Easy to reuse

Tips for Building Your First Skill

If you want to get started, follow these guidelines:

  • Start with your most common workflow

  • Upload your real template

  • Use plan mode on the first run

  • Expect to iterate

  • Test across multiple deals

Iteration is key. Most skills improve after a few refinements.

Is It Worth Paying for a Shortcut?

It depends on your workflow.

Use Claude If:

  • You run occasional analysis

  • You don’t need standardization

  • You prefer flexibility

Use Shortcut If:

  • You process multiple deals weekly

  • You need consistent outputs

  • You want repeatable workflows

Time & Efficiency Impact

Factor Traditional AI Use Skills Workflow
Setup Time Repeated One-time
Output Consistency Low High
Rework Needed High Low
Workflow Speed Moderate Fast

FAQs Regarding AI Underwriting Skills Workflow

What are AI skills in underwriting?

They are repeatable workflows that automate deal analysis.

  • Store templates

  • Standardize outputs

  • Reduce manual work

  • Improve consistency

You can explore workflow automation concepts at Zapier (https://zapier.com/).

Why is consistency important in underwriting?

Consistency improves clarity and decision-making.

  • Easier comparisons

  • Better reporting

  • Clear structure

  • Faster reviews

Notion (https://www.notion.so/) shares insights on structured workflows.

Can skills replace manual underwriting?

No, they support but do not replace analysis.

  • Automate repetitive tasks

  • Improve speed

  • Require oversight

  • Enhance accuracy

IBM (https://www.ibm.com/) discusses AI in business workflows.

How long does it take to build a skill?

Most skills can be built in under 15 minutes.

  • Define trigger

  • Upload template

  • Set rules

  • Test output

Automation Anywhere (https://www.automationanywhere.com/) explains workflow automation basics.

Do skills work across different deals?

Yes, if tested and refined properly.

  • Handle multiple inputs

  • Adapt to deal types

  • Require testing

  • Improve over time

UiPath (https://www.uipath.com/) covers scalable automation systems.

What are the biggest limitations?

Skills may need iteration to perform well.

  • Not perfect initially

  • May miss edge cases

  • Requires refinement

  • Needs validation

For AI limitations, see DeepMind (https://deepmind.com/).

Can skills improve team workflows?

Yes, they make processes more consistent.

  • Standard templates

  • Faster onboarding

  • Better collaboration

  • Clear outputs

Slack (https://slack.com/) discusses team productivity tools.

Is this useful for small teams?

Yes, especially for lean teams handling multiple deals.

  • Saves time

  • Reduces errors

  • Improves output

  • Scales easily

Airtable (https://www.airtable.com/) provides workflow tools for teams.

Build Consistency Into Your Workflow

Join the AI for CRE Collective, where 600+ CRE professionals are using tools like this to standardize underwriting, eliminate messy outputs, and scale their deal pipeline with confidence.

Get access to real templates, step-by-step setups, and the 12-month Perplexity Pro access—and start turning inconsistent AI results into repeatable, reliable workflows.

Leave a Reply

Your email address will not be published. Required fields are marked *