Find out which parts of your business
are worth automating.

In two to three weeks we watch how your team really works, count what repeats, check what your systems can connect to, and give you a plan that says what to automate first. We have been building business software since 2012.

This is for you if

Services

Every piece of work included, in order

Most of this happens at your desks, not in a meeting room. What people describe in a meeting is rarely what they actually do.

01

Write down how the work really happens

Every step, including the workarounds that never made it into the written procedure. We draw it in a standard format, so a developer can build from it later without having to guess what a paragraph meant.

02

Measure how much of it there is

How often each task happens, how long it takes, and how many people touch it. Taken from your own system records where they exist, not guessed in a meeting. People guess wrong in both directions.

03

Check whether your systems can be connected

For each system we check whether there is a proper way in, or whether we would have to work around it. Nothing else on this list changes the cost as much as this does.

04

Test how well your documents can be read

We run your real paperwork through several tools and compare what each one reads correctly. A clean invoice, a delivery note and a scan with handwriting on it give very different results, so we test it instead of assuming.

05

Work out what each process is worth

How much time it takes now, what the mistakes cost when they happen, and roughly what it would cost to fix. This is what lets you compare two processes that have nothing else in common.

06

Say whether to build or buy

Some of this is already solved by software you can buy, or by a part of the system you already own and do not use. We will say so, even though it means less work for us.

07

Check which rules apply to you

What your auditor will ask, what data is allowed to leave your systems, and what has to be agreed in writing before anything is built. Much cheaper to sort out now than after it is built.

08

Put them in order

Which one to do first, and why. Usually not the biggest one. The best first project is a cheap one that proves it works, which makes the next two faster and safer.

09

Say what to leave alone

Some processes would cost more to automate than they save. Some happen too rarely to bother. In others the real problem is the process itself. This is usually about a third of what we look at.

Thirty minutes is usually enough for us to tell you whether this is worth doing at all.

Our stack

The tools, standards and methods we use

You do not need to know any of these. They are listed so you can see we use the same standards and tools as the rest of the industry, not our own invented method.

Writing down the processWe use the standard diagram formats that most workflow and ERP tools can already read, so your rules are not locked to us
BPMN 2.0DMN decision tablesValue stream mapping
Checking what really happensThese tools read the time stamps your systems already record, so we are not relying on what people remember
Process miningCelonisPower Automate Process MiningApromoreTask mining
The numbers we measureThe standard set, so two completely different processes can be compared fairly
Cycle timeTouch timeFirst-pass yieldRework rateException rate
Reading documentsWe test these on your own paperwork, because accuracy changes a lot with the type of document
Azure AI Document IntelligenceGoogle Document AIAWS TextractLLM extraction with JSON schemaOCR quality assessment
Connecting your systemsWhat is already available decides the cost. Software that clicks through screens is a last resort, not a first choice
RESTSOAPGraphQLOAuth 2.0Database viewsScheduled exportsRPA fallback
Ready-made tools we check firstWe look for something you can buy before we suggest building anything
UiPathAutomation AnywherePower Automaten8nERPNext and Odoo modules
Rules you have to followWhich ones apply depends on who you sell to and who audits you. We tell you which
EU AI Act risk tiersISO/IEC 42001NIST AI RMFGDPR and data residencyData processing agreements
Working out the real costIncluding what it costs to keep running, which most quotes leave out
Three-year total cost of ownershipPayback periodPer-document run cost

The plan

What the plan contains

Automation planPrepared for your operations team
  1. The process maps in BPMN, and your decision rules as tables
  2. Measured volumes per process, from your own records
  3. The ranked list, and why it is in that order
  4. A note on each system: how it connects and what that costs
  5. Document accuracy results from testing your real paperwork
  6. The first project we recommend, described in enough detail to start
  7. What we would not touch, and why
  8. The rules that apply to you, and what they mean in practice

Written to be used, not filed. Every process is named, every number comes from your business, and the first project is described in enough detail for a developer to start on it.

It works without us. The plan is yours. Some clients have used it to brief their own team, or to get quotes from other companies. That is fine, and we price the work on that basis.

It says no as often as yes. The list of what not to automate usually saves the most money.

Example

What we found for one distribution client

Example

A distribution company asked us to look at their back office. We found eleven processes that repeat. Four were worth automating, three needed the process fixed first, and four were not worth touching.

Supplier invoices came first. Not because it was the biggest job, but because it happened about 310 times a month, the rules for handling a mismatch were already written down, and their ERP had a proper way in. Together that made it the cheapest way to prove the idea works before spending on anything bigger.

This is an example, not a promise. Your processes and numbers will be different, and your plan will have your own figures in it.

FAQ

Questions we get asked before starting

How long does it take?

Two to three weeks for most companies. Longer if the work is spread over several sites. If you already know which process you want to fix, you may not need this step at all.

What do you need from us?

A few days with the people who do the work, and one senior person who can decide what should happen in the difficult cases. You do not need to put a project team together.

Do you have to be on site?

For most of it, yes. Watching the work is what finds the problems. We can do parts remotely where your system keeps good records, but the watching is the part that matters.

What if the answer is that nothing is worth automating?

Then the plan says so, and you have saved far more than it cost. This does happen, and it is a good result, not a failed project.

Are we committed to building anything with you?

No. The plan is yours and you can take it anywhere. Some clients use it to brief their own team, or to compare quotes from other companies.

We are not a manufacturer. Does this apply to us?

Yes. We see the most repeated paperwork in manufacturing, logistics and warehousing, but every back office has some. The method is the same in any industry.

Next step

Tell us what your team does over and over.

Tell us one process, roughly how often it happens, and what goes wrong. On the call we will say whether this study is worth doing, or whether you should skip it and go straight to building.

Ahmedabad, India. We work with teams in the US, UK, Europe, Singapore and the Gulf, and we are used to the time difference.