MN Consulting / Workflow automation

Workflow automation

Map the process and the data before you add AI

For small and mid-sized service businesses: accounting firms, insurance brokers and other professional-services firms. AI tools are easy to buy. The harder part is knowing how the work really runs: who does what, in which system, and where the data comes from. Without that, AI speeds up a process nobody has written down.

  • Agreed in writing first
  • No lock-in
  • Uses the systems you already pay for
Sound familiar?
  • Teams re-keying data between systems and building reports by hand.
  • AI tools bought before anyone has written down how the work really runs.
  • Each new client set up by hand in each system.
Common places to start
  • Client onboarding: collecting documents and setting the client up in each system.
  • Data re-keyed between the practice management system, the accounting package and spreadsheets.
  • Month-end and management reports built by hand.
  • Time, billing and work-in-progress follow-up.
  • Engagement letters and other standard documents.
  • Insurance renewals chased manually, by email and phone.
The order I work in

Structure first, then your systems, then AI where it pays

  1. 1 · Structure

    Map the process and the data

    I map the process as it really runs, not as the procedure says, including the handoffs and the rework. For each step I write down what data it needs, which system it sits in, who owns it and how it moves.

  2. 2 · Existing systems

    Use what you already pay for

    Most firms already have software that can do more: practice management, document management, the accounting package and the office suite. I automate the re-keying and the manual reports with those first, so nothing new is bought until it is needed.

  3. 3 · AI where it pays

    Add AI to the steps that justify it

    With the process and the data structured, I use AI where it saves time and a person can check the output, such as sorting incoming documents or drafting a first version. Where a rule or an existing system does the job, I use that instead.

How it works

Review, fix, hand over

Typical problems I fix
  • Month-end and management reports built in spreadsheets.
  • Client onboarding run by email, and each new client set up by hand in each system.
  • Data re-keyed between systems, and renewals chased manually.

Background: process and operating-model review roles in large professional-services firms.

  1. 1
    ReviewThe process mapped as it really runs, and the data each step needs.
  2. 2
    FixI automate it with the systems you already have, and AI where it pays.
  3. 3
    Hand overDocumented, tested and owned by your team.
How an engagement runs
First 30 days
The process mapped as it really runs, and the first workflow agreed for build.
What I do
I build the automation, test it and hand it over documented.
Done when
The workflow runs, documented and owned by your team.
How it is priced
A monthly engagement. The outcome, the first 30 days and the fee are agreed in a statement of work before anything starts. After the first month it runs month to month.
Who you will work with

Shehan Mathew

I run MN Consulting. I worked in process and business-improvement roles at Linklaters and EY, both professional-services firms. LinkedIn

Questions

Should we start with AI?

Not usually. Start by mapping the process and the data it depends on. Once that is written down, it is clear which steps your existing systems can automate, and where AI would save time.

Do we need to buy new software?

Not at the start. I work with the systems you already have first, and only suggest something new where they cannot do the job.

What does the first month cover?

The process mapped as it really runs, and the first workflow agreed for build. The outcome, the first 30 days and the fee are agreed in a statement of work before anything starts. After the first month it runs month to month.

Re-keying data or building reports by hand?

If you want to start in the right order, I am happy to talk it through.

Book a 30-minute call
Email
info@mnconsulting.io