Handled by Morning

Handled by Morning / Private AI for business

Private AI for your business, set up on your own data.

Your staff are already pasting company numbers into public chat tools. I set up one private assistant in accounts you own, connected to your real records, with written limits on what it may see and do.

  • Setup, fixed$6,500
  • Then$750 a month
  • Working infour weeks
You ask

Which customers are more than 30 days late, and how much?

Four customers owe $41,150 that is more than 30 days late. One of them is over half of it.

CustomerDays lateAmount
Harbor Supply62$22,400
Delmar Builders47$9,850
Kessler & Ruiz38$5,600
Pacific Ridge HOA33$3,300

Harbor Supply has paid late on five of its last six invoices.

Where this came from
  • QuickBooks, invoices
  • Updated 5:40 a.m.
A sample. The company and figures are made up.

Reads from what you already use

  • QuickBooks
  • Xero
  • HubSpot
  • Salesforce
  • Excel
  • Google Sheets
  • Google Drive
  • SharePoint
  • Snowflake
  • Power BI

Something else? I check before I quote.

What happens now

  • Staff pasting customer and financial data into public tools
  • No record of what was shared or where it went
  • Answers that sound right and cannot be checked
  • Nobody sure what the AI vendor does with your data

What gets set up

  • One assistant in accounts you own, on plans that do not train on your data
  • Connected read-only to your books, sales and documents
  • Rules for who can ask about what, agreed in writing
  • Every answer shows its sources, and a log of what was asked

And every morning, the page you would have asked for

Most days you should not have to ask anything. One page lands in your inbox before you open, with yesterday next to a normal day.

The line at the bottom reads the numbers for you and points at the one thing worth a conversation.

Reporting is where I come from. If you need more than one page, I also build dashboards, forecasts and multi-location reporting on their own.

Morning report
TuesdaySent 5:40 a.m.
YesterdayUsual Tuesday
Sales$18,420$16,900
Orders4138
Gross margin31.2%34.0%
Cash collected$9,300$12,100
Overdue, all customers$41,150$33,800
Open quotes$126,000$118,000

Margin fell almost three points. Two large orders went out at last year's prices, and both were for the same customer.

Sample. The names and numbers are made up to show the layout.

Supplier bill 8841Sample
  1. A bill arrives by email from Harbor Supply.

    Inbox
  2. Read: supplier, $2,318.40, due in 30 days, nine line items.

    Automatic
  3. Matched to purchase order 1127. One line is $84 over what was ordered.

    Automatic
  4. Drafted in your books with the difference noted.

    Automatic
  5. Nothing is posted or paid.

    Waiting for your OK

Paperwork read, checked and drafted before anyone opens it

Once your systems are connected, the same setup can do the reading and typing that eats an office's week. You pick two jobs to start.

  • Supplier bills into the books
  • Quotes drafted from past ones
  • Contracts searched and summarised
  • Customer forms entered
  • Month-end checks
  • The weekly report, written

It drafts and flags. Anything that touches money or goes to a customer waits for a person.

Private means your data stays yours

The fair worry is that company numbers end up somewhere they should not, or that staff paste them into a public chat tool because nothing better exists. This replaces that with one place that is set up properly.

Before anything goes live, the limits below are written down and agreed with you.

If your data cannot leave the building at all, it can run on a machine in your office. That takes different hardware, and I will tell you honestly whether you need it.

  • Where it livesIn accounts you own. Your data is not copied into mine.
  • TrainingOnly on business plans whose terms say your data is not used to train AI models. I show you the terms.
  • Who can ask whatYou decide. Payroll and bank detail stay with the people you name.
  • SourcesEvery answer shows where it came from, so you can check it.
  • When it does not knowIt says so. It does not guess a number.
  • Your recordsIt reads. It never changes your books, pays or sends anything on its own.
  • Off switchOne setting turns it off, and the business runs the way it did before.

Your four options, side by side

Spreadsheets, as nowChatGPT for the staffHire an analystThis setup
To start$0A fee per personRecruiting, then a salary$6,500
Each monthSomeone's Monday morningA fee per personA full salary$750, plus tools
Knows your numbersWhoever built the sheet doesOnly what staff paste inYes, after a few monthsYes, connected to your systems
Where your data goesStays putWherever staff paste itStays putStays in your accounts
If someone leavesThe sheet breaksNothing was builtThe knowledge leaves tooIt keeps running, and you own it

Hiring an analyst is the right call for some companies. On the free call I will say so if you are one of them.

How the four weeks go

Week one

Connect and check

Read-only access to your systems. I pull a year of history and compare my totals with yours until they match. On Friday you see what your data can and cannot answer.

Weeks two and three

Define and build

We agree what each number means and who may see it. Then the assistant, the morning report and your two paperwork jobs are built on top.

Week four

Run it for real

Your team uses it while I read every answer it gives. We fix what is wrong or unclear, I teach your people, and from then on it is yours.

A setup price, then a monthly one

$6,500

then $750 a monthcancel any month, keep everything

Setup is fixed and takes four weeks. The monthly price covers watching it, fixing it and adding to it as the business changes. The AI and data tools bill you directly, and I give you that number before we start.

  • Up to four systems connected read-only
  • Your numbers defined and checked against your own totals
  • A private assistant for you and your staff answers with sources
  • One morning report emailed daily
  • Two paperwork jobs you pick them
  • Written limits and access rules agreed with you
  • Training for your team recorded

If what I show you at the end of week one does not convince you, you pay nothing and keep the findings.

Book a free look
Peter Skotte

You work with the people who build it

I'm Peter Skotte, a data analyst in San Diego. I have spent my career building reporting and forecasts for restaurant, retail and finance operators, and assistants that answer plain-language questions from checked numbers.

Wyatt Milne works with me on dashboards and reporting. He has built business reporting since 2019 for a mortgage lender, a multi-business holding company and a national restaurant brand, and holds Microsoft and Domo certifications in it.

For other parts that need it, I bring in more people I have worked with in enterprise analytics, software and data architecture. I stay the lead and your one contact.

Based in
San Diego, California
Company
Handled by Morning
Meets
In person across San Diego County, by video elsewhere

Book a free look at your numbers

No pitch deck and no jargon. Here is what happens after you send this.

  1. I call or email you within one business day.
  2. You show me where your numbers live, on a 30 minute call or in person in San Diego County.
  3. I tell you what your data could answer today, what it could not, and what it would cost. If it is too early for you, I say so.

Or call me directly.
(858) 254-0491

Where do your numbers live?
Your details go to me and nowhere else.
Call (858) 254-0491Book a free look