Sample AI Workflow Day deliverables: what a team takes home.

People ask what the day actually produces. This is the honest way to show it: a full set of deliverables for a fictional company, Get Clear Example Company, laid out exactly as a real team would receive them, with every number left blank for a real team to measure.

About this sample

Illustrative sample. Get Clear Example Company is a fictional company invented to show what an AI Workflow Day leaves behind. It is not a client, and nothing on this page is a measured outcome. The people, tools and workflow are made up. No results are claimed; every measurement field is left for a real team to fill in.

The workshop in brief

• Workflow selected: Turning technician job notes into customer-ready service reports and follow-up quotes
• Why this one: Every service call produced one. Three people touched each report. Quotes for recommended repairs were often sent days late or not at all.
• In the room: Owner, service manager, two technicians, dispatcher, office administrator, bookkeeper (7 people)
• Tools already in use: Text messages and photos from the field, a job log spreadsheet, email, accounting software, a shared price sheet
• Format: One day on site, preparation the two weeks before, follow-up session 30 days after
• Workflow owner: Service manager

The workflow before the day

• 1. Technician: Texts job notes and photos to the dispatcher from the truck
• 2. Dispatcher: Copies the notes into the job log spreadsheet
• 3. Office administrator: Writes the service report from the notes, texting the technician when something is unclear
• 4. Service manager: Reviews the report for technical accuracy and tone, edits, sends it back
• 5. Office administrator: Builds a quote for the recommended repairs from the price sheet, if there is time
• 6. Service manager: Approves the quote
• 7. Office administrator: Emails the report and quote to the customer, logs both in the spreadsheet and the accounting software

The workflow after the day

• 1. Technician: Records a voice note and photos in the job form before leaving the site, answering five fixed prompts (Same effort as the old text, better structure)
• 2. AI assistant: Drafts the service report and the recommended-repairs section from the notes, using the SOP template and the company's plain-language style
• 3. Technician: Reads the draft on the phone, corrects technical details, marks it ready (Human approval point 1)
• 4. Office administrator: Pulls quote lines for the approved recommendations from the price sheet; the assistant drafts the customer note that goes with them
• 5. Service manager: Approves the report and the quote in one pass (Human approval point 2)
• 6. Office administrator: Sends both to the customer and logs them

Safe-use rules the team agreed

• Use the company AI account only. Personal accounts are not used for company work, even for a quick draft.
• Customer names, site addresses and equipment details may go into the assistant. Payment details, credit applications, and anything from another customer's file may not.
• The assistant drafts. A person who knows the job approves. Nothing reaches a customer without a named approver.
• Prices, part numbers, warranty terms and safety statements come from the price sheet, the manufacturer or the manager, never from the assistant.
• If the assistant states something that is not in the source notes, delete it and tell the service manager so the prompt can be adjusted.
• Photos of customer premises stay in the job system. They are described to the assistant in words when needed, not uploaded to it.
• Keep the prompt set in the shared folder. Edits go through the workflow owner so everyone is working from the same version.
• These are Get Clear Example Company's operating rules for this example, the kind a team agrees on the Workflow Day. They are not legal, privacy or security advice. A real team should have its accountant and insurer review rules like these.

The ranked opportunity map

• 1. Service reports and follow-up quotes. Value high, risk low, difficulty low. Built on the Workflow Day in this example. The first workflow.
• 2. Maintenance agreement renewal reminders. Value high, risk low, difficulty medium. First 30 days. Same prompt style, new template.
• 3. Parts order summaries from supplier emails. Value medium, risk low, difficulty low. First 30 days. Bookkeeper owns it.
• 4. First-draft replies to online reviews. Value medium, risk medium, difficulty low. Days 31 to 60. Owner approves every reply.
• 5. Daily crew briefing from dispatch notes. Value medium, risk low, difficulty medium. Days 61 to 90. Needs the dispatch notes in one place first.
• 6. Estimating full system replacements. Value high, risk high, difficulty high. Not yet. Pricing data is in three places. Revisit after the 90-day review; likely a custom tool.

Measurement fields

• Hours per week spent writing reports and quotes: baseline measure before workshop, then at 30 and 90 days.
• Days from job completion to report sent: baseline measure before workshop, then at 30 and 90 days.
• Follow-up quotes sent per week: baseline measure before workshop, then at 30 and 90 days.
• Quotes sent within one business day of the job: baseline measure before workshop, then at 30 and 90 days.
• Customer corrections or complaints about reports: baseline measure before workshop, then at 30 and 90 days.

Services · Work · Guides · About · Contact