Level 1 credential · 8 clock hours
AI Automation BuilderBuild it. Guard it. Run it.
Learners direct an AI coding agent to build an automation for a real bottleneck at work, put a person in charge of what leaves the building, test it until ten of ten cases pass, and leave it running.
Hands-on training using Claude Code
The credential at a glance
- Credential
- AI Automation Builder Certificate
- Type
- Certificate
- Clock hours
- 8
- Format
- Four 2-hour build sessions, one a week, in person
- Price
- $1,000 a seat, invoiced after the last session
- Class size
- 8 to 15 people
- Assessment
- Rubric on the work itself, a live demo and a 20-question quiz
- Issued by
- Paglow Automations (Ohio AI Solutions LLC), Lima, Ohio
- Before the first session
- Comfortable with email, files and spreadsheets. No programming experience needed.
- A Windows or Mac computer the learner can install the Claude desktop app on.
- A paid Claude seat that includes Claude Code, opened by the employer before session 1.
- The opening lessons of Anthropic's free Claude Code 101 course, done before session 1.
- One real bottleneck, picked with a manager in a 15-minute kickoff.
What a holder can do
TechCred pays for applied skills in developing, building, configuring or operating technology. Every competency below ends in something the learner built or ran, and an assessor checks the evidence.
Build
Make the thing that does the work.
Write an automation's job description — inputs, steps, output, approval point, success test — as instructions an AI coding agent follows.
Evidence: The project's CLAUDE.md
Direct an AI coding agent to build a multi-step automation that reads a business input, runs an AI step with a fixed output format, and writes a draft output.
Evidence: Automation running end to end on test data
Configure
Set the rules it runs under.
Sort business data into green, yellow and red, and set what an AI tool may see.
Evidence: Data rules sheet signed off by the manager
Package the automation as a one-word command a coworker can run.
Evidence: Skill file and a coworker's run
Configure permission rules, a human approval step before anything leaves the business, and a hook that holds risky actions.
Evidence: Settings file and written approval rules
Operate
Run it, test it, keep it running.
Test the automation against 10 real, sanitized cases with expected results, and fix it until all 10 pass.
Evidence: Test log at 10 of 10
Run the automation on live work — on demand, or on a schedule once it has passed its tests — with a run log and a runbook.
Evidence: Live run, run log and one-page runbook
Analyze
Measure the work before and after.
Map a business process step by step and measure its time and cost per week.
Evidence: Process map and baseline worksheet
Measure the before-and-after time and cost, and demonstrate the automation in five minutes.
Evidence: Before-and-after numbers and a live demo
Two hours each. Something built every time.
One session a week, in person. Each one ends with an artifact the learner keeps and the assessor checks.
Session 1 · 2 hours
Map it, spec it
A measured bottleneck and a written job description for the automation.
- A step-by-step map of one real process
- A time-and-cost baseline for a normal week
- Green, yellow and red rules for the data involved
- The automation's job description, saved as the project's CLAUDE.md
- A first Claude Code session that reads sanitized samples and proposes a plan
Proof: Scored map, baseline and spec
Session 2 · 2 hours
Build it
The automation running end to end on test data.
- Read the input: an export, a folder, a shared sheet or a mailbox
- An AI step that returns the same fields every time
- Write a draft output: a sheet row, a document or a reply draft
- Package it as a one-word command a coworker can run
Proof: Automation running on test data
Session 3 · 2 hours
Guard it, test it
Guardrails in place and 10 of 10 test cases passing.
- Permission rules: what it may read, change and never touch
- An approval step before anything leaves the business
- A hook that turns any send into a question for a person
- Handling for missing fields, bad files and odd requests
- Ten real, sanitized cases with expected results, fixed until all pass
Proof: Test log, approval rules and settings file
Session 4 · 2 hours
Run it, prove it
Live in the business, documented, measured and demonstrated.
- Switch to live work, on demand or on a schedule at the slowest pace the job allows
- A run log anyone can read
- A one-page runbook: how to run it, check it, fix it and turn it off
- Before-and-after numbers
- A five-minute demo and the quiz
Proof: Live demo, runbook, numbers and quiz
How it's earned
The credential is earned on the work, not on attendance alone. An assessor scores the learner's own artifacts against six criteria.
Proficient or better on all six rubric criteria, 80% or more on the 20-question quiz, and all 8 hours attended.
Criterion 1
Map and baseline
Proficient means
Every step of the process is mapped, and the weekly time and cost come from counted examples, not guesses.
Criterion 2
Data rules
Proficient means
Each data type is classed green, yellow or red; no red data reaches the AI step; the manager signed off.
Criterion 3
Build
Proficient means
The automation runs end to end from a one-word command and returns the same output fields every time.
Criterion 4
Guardrails
Proficient means
Nothing leaves the business without a person's approval; risky actions are held by rule and hook; failures stop cleanly with a message.
Criterion 5
Testing
Proficient means
Ten real, sanitized cases with expected results are logged, every failure has a recorded fix, and all ten pass.
Criterion 6
Operation and results
Proficient means
It runs on live work, the runbook lets a coworker operate it, and the before-and-after numbers are measured.
- Not yet
- Developing
- Proficient
- Exemplary
Every criterion has to reach Proficient.
A missed session is made up from the recording plus a 30-minute build check before the next session. Each certificate carries a verification code anyone can check at paglow.com/training/verify.
Your business owns the accounts.
They're opened by your business, billed by the software companies, and never part of the training invoice. Owning them is what keeps the work running after class.
Claude (Pro or Team plan)
One seat per learner, with Claude Code included. The learner builds in the Claude desktop app.
Pro is $20 a month; Team is $25 a seat a month ($20 billed yearly), two seats or more. Billed by Anthropic.
Software subscriptions are not reimbursable under TechCred, so the seat is yours directly and never on the training invoice. On Team, business data is not used to train models by default; on Pro, the learner turns that setting off in session 1.
Paglow isn't affiliated with Anthropic or Vapi. Claude and Claude Code are Anthropic's products; Vapi is Vapi's.
What never goes into an AI tool
In session 1 every learner sorts their data into green, yellow and red. Red never goes into an AI tool. These are the red lines each industry starts from.
No patient information goes into any AI tool in class, ever.
Manufacturing
Red
- Export-controlled drawings or data (ITAR, EAR)
- Employee personal and medical records
- Customer data a contract keeps confidential
Healthcare administration
Red
- Anything that identifies a patient or their care (all PHI), including a name tied to a visit request
- Insurance member IDs and claims
- Staff licenses and personnel files
Trades and home services
Red
- Payment card numbers
- Gate codes and alarm codes
- Employee personal records
Professional offices
Red
- Social Security, bank and account numbers
- Privileged or confidential client material, unless office policy and a business-plan account allow it
- Tax returns and financial statements outside an approved account
What people build, by industry
The skills are the same in every class. These are the kinds of automations learners pick from in the manager kickoff.
Manufacturing
Plants, machine shops and fabricators: estimating, purchasing, maintenance and the front office.
- Quote requests, sortedRFQ emails become a quote-prep sheet — part, quantity, material, due date — plus a reply draft asking for anything missing.
- Maintenance requests, rankedFree-text requests from the floor become a work-order log, ranked by safety and downtime.
- Supplier delays, caughtSupplier emails become an expedite list: the new date and every order it touches.
- Shift notes, summarizedEnd-of-shift notes become one page the next lead reads in a minute.
Healthcare administration
Clinics, dental, therapy and home-health offices — the front desk and back office, never clinical work. Anything that touches a patient needs a HIPAA-covered setup.
- Payer bulletins, explainedInsurance payer bulletins become a one-page summary of what changed for the front desk.
- Forms and directions answer bankThe office's approved facts become a bank of reply drafts staff paste into answers about forms, hours and directions.
- Shift coverageStaff swap requests become a coverage sheet with every gap flagged.
- Supply reordersWeekly supply counts become a reorder list against par levels.
Trades and home services
HVAC, plumbing, electrical, roofing, cleaning and contractors — the office behind the trucks.
- Leads in, sortedWeb-form and email leads become a lead sheet with job type, address and urgency, plus a reply draft.
- Estimates followed upOpen estimates older than a week get follow-up drafts in the owner's own words.
- Done jobs, billedCompleted job notes become invoice line items and a review-request draft.
- Permits and inspectionsInspection emails become a tracker with dates, results and what's next.
Professional offices
Law, accounting, insurance and real estate offices — intake, documents and follow-up.
- Intake to summaryA new-client questionnaire becomes an intake summary and a list of names for the conflict check.
- Document chaserA missing-documents list becomes reminder drafts for each client, in the office's voice.
- Quote requests, preppedInsurance quote requests become a quote-prep sheet with the missing details flagged.
- Listings with a fair-housing checkProperty details become a listing draft, checked against fair-housing wording rules before anyone sees it.
Plan a class with this credential
Classes start in 2027. Tell me about your team and I'll reply myself with dates, a plan and the TechCred steps.