LeadLoup
STANDBY
Book a callFR

Process automation

Monday morning, the follow-ups have gone out, the records are current, the log is written.

You write the rule once: which delay, which condition, which exception. The agent applies it from then on without supervision, and records every action in a dated log. When the work calls for judging a situation rather than applying a rule, what you need is a custom agent.

Book a call

19.2%

of Canadian businesses used AI to produce their goods or deliver their services in the past 12 months. Two years earlier it was 6.1%, so the share has tripled in two years.

Statistics Canada, Canadian Survey on Business Conditions, Q2 2026

One day, hour by hour

Demo

Typical hourly profile for a service business. Demonstration values, not operating data.

Where automation fits, and where it does not.

Yes, if

  • The same tasks come back every week and always land on the same person.
  • Quotes sit without a follow-up because the week got busy.
  • Somebody is still copying data from one tool into another by hand.

No, if

  • Your way of working changes every month. There is no stable process to automate yet.
  • You are looking to replace a team. The agent runs the tasks; the decisions stay with you.
  • Nothing is written down anywhere. Start with the strategy work: we map before we automate.

How a forgotten task becomes a task that runs

  1. The trigger gets named

    A quote with no answer after three days, a file left open, an invoice coming due: we set with you the exact moments a follow-up has to go out, from your real rules rather than a template imported from another trade.

  2. It runs with nobody touching it

    At the appointed moment it sends the follow-up email, updates the status in your tracking tool, or warns the right person at your end. Nothing manual is needed for the action to go out on time, including on a Friday night or during vacation.

  3. The result is logged

    Every follow-up sent and every update made is recorded and dated in a log you can open at any time. You see what was done without asking, and you can tighten a rule when a special case starts coming back too often to stay an exception.

What you get

10 deliverables

  1. The task mapThe ones worth doing, ranked by hours saved.
  2. Your trigger rulesDelays, conditions, exceptions, written in your own words.
  3. Your tools connectedEmail, CRM, spreadsheet, invoicing, with no new subscription.
  4. The message approved up frontYou sign off on the tone before anything goes out.
  5. The action logEvery action dated, with its trigger, open at any time.
  6. The first month reportHow many follow-ups went out, how many got an answer.
  7. The rules documentedSo somebody else can change them without starting over.
  8. A run on real casesBefore the trigger goes live for good.
  9. An adjustment sessionTo tighten a rule that turned out too loose or too strict.
  10. The plan for what followsThe next tasks to automate, once the first ones hold up.

What people ask us

How is this different from a custom AI agent?
Automation applies a rule you wrote: after three days with no answer, this email goes out. It does exactly what was planned, it is predictable, and it costs less. An agent reads a situation and decides what to do with it, which is what you want when the right action depends on what the client answered. A fixed rule for mechanical work, an agent when judgment is needed. Most companies end up with both.
Which tasks are worth automating?
The ones that come back often, follow a clear rule, and cost time without calling for judgment every time. A quote that follows up after three days of silence is a textbook case. A contract negotiation is not, because it calls for judgment case by case.
What if the task has an exception nobody saw coming?
The agent applies the rules it was given and flags cases outside the frame instead of guessing an action. Those cases then feed the rules, so the same exception is covered on its own the next time it shows up.
Do we have to change the tools we use today?
No. The agent connects to what you already have: email, CRM, spreadsheet, invoicing software. Nothing to migrate, nothing for your team to relearn, and no extra dashboard to watch on top of the ones you already open.
How long before the automation is running?
Scope sets the timeline. A single isolated task goes live in a few days; several linked processes take longer to map properly. The first meeting is where a real schedule for your case gets set, not a generic delay copied from another engagement.
What happens when the volume of work jumps all at once?
The agent handles every trigger the same way, whether there are five or five hundred in a day. Volume changes neither the reliability of the run nor the delay before a follow-up goes out, unlike a person who is buried.
What does this change on my website?
At one end, almost always: your site form is a trigger like any other. A request coming through it can land in your CRM, acknowledge receipt and start the follow-up without anybody retyping a thing. We do not rebuild your site for that, we connect to the one you have. If the site itself is what holds you back, that is a different job.
We are looking for an AI assistant for the company. Is that what you do?
Partly. An assistant waits to be asked; what gets installed here goes off on its own, at the appointed moment, with nobody thinking about it. When it also has to read the situation and decide, what you need is a custom agent. Most companies end up with both.

See which tasks in your shop would follow up on their own.

Book a call

20 minutes with David, founder of LeadLoup, no commitment. You leave with a concrete list of the tasks that can be automated first at your place, ranked by how easy they are to set up.