Services
Different realities deserve different technological answers.
Once we understand how a business works, the answer can be software, a connection, an automation, useful information or a better customer experience. Sometimes it is nothing new.
01 A process the business cannot see.
Make it visible.
What it looks like
Work lives on a whiteboard, in a chat and in someone’s memory. Nobody sees the whole until something is late.
How we would approach it
- 01
Give every unit of work a record and a state, in the states the team already uses, not a template’s.
- 02
Show each role only what it acts on: packing an order and deciding about it need different views of the same record.
- 01
02 The same information, kept in several places.
Connect it.
What it looks like
A price or a stock count lives in three tools and agrees in none, so someone reconciles them by hand.
How we would approach it
- 01
Decide which system owns each piece of information before connecting anything. A connection without an owner copies the disagreement faster.
- 02
Use each tool’s API where one exists and a scheduled sync where it does not. A person shouldn’t have to be the permanent connector between systems.
Where we hold back
Sometimes the fix is fewer places, not more connections: retiring a spreadsheet can do more than integrating it.
In our work
Hortiluma’s counter, electronic invoicing, orders and online store run on one system, Farol, in production. Hortiluma belongs to the same family behind Dimerso.
- 01
03 A handoff that consumes someone’s attention.
Automate what should be automatic.
What it looks like
An order arrives in one place and someone types it again in another. The work is simple; the attention it takes is not.
How we would approach it
- 01
Separate the mechanical part of a handoff from the judgment in it. Automate the first; give the second a better view.
- 02
Trigger from the event itself, an order placed or an invoice issued, not from someone noticing it.
- 03
Where the input is loose language, such as an emailed order, a model can structure it, with a person confirming before anything is sent.
- 01
04 An operation bent around generic software.
Build what belongs.
What it looks like
The team works around the tool: exports, side spreadsheets, steps done outside because the software has no place for them.
Where we hold back
If a standard product fits with sensible configuration, we recommend it. Custom software is a commitment, and it has to earn it.
How we would approach it
- 01
Model the data on the operation’s own vocabulary and on how the work actually moves.
- 02
Build the smallest complete version that can run a real day, then extend it where the work shows it needs to go.
In our work
Farol began this way: a sale, an invoice, a stock count and a collection kept in step by hand, in business after business in Costa Rica, until the pattern became its own system.
Farol- 01
05 Decisions made without usable evidence.
Make the information useful.
What it looks like
The data exists, but a decision still waits for someone to build a spreadsheet, or is made by instinct.
How we would approach it
- 01
Start from the decision, not the dashboard: which question, who asks it, how often, and what the answer would change.
- 02
Fix where the numbers come from before charting them. A clean view of untrusted data only hides the problem.
- 03
Use applied intelligence where a probable answer helps, such as anticipating demand, and plain queries where the answer must be exact.
Where we hold back
A decision made twice a year may need a well-kept report, not a live view or a model.
- 01
06 A customer experience disconnected from the business behind it.
Redesign both sides.
What it looks like
The website promises one thing and the operation delivers another: a booking is received, and nothing behind the page knows it exists.
How we would approach it
- 01
Design the customer’s path and its operational counterpart together: every promise on screen needs a place in the operation that keeps it.
- 02
Let the experience read the records the team works from, so availability, status and prices are true when a customer sees them.
Where we hold back
If the operation cannot keep a promise yet, it does not go on the page.
In our work
BeautyBox: a live public booking experience built around its services and locations.
Cases- 01
The material
Rarely one technology alone.
Most responses draw on more than one material. One draws on none.
| Situation | 01Purpose-built software | 02Operational systems | 03Automation | 04Integration | 05Information & analytics | 06Applied intelligence | 07Digital experience |
|---|---|---|---|---|---|---|---|
| A process the business cannot see. | Operational systems | Information & analytics | |||||
| The same information, kept in several places. | Integration | Information & analytics | |||||
| A handoff that consumes someone’s attention. | Automation | Integration | Applied intelligence | ||||
| An operation bent around generic software. | Purpose-built software | Operational systems | Integration | ||||
| Decisions made without usable evidence. | Information & analytics | Applied intelligence | |||||
| A customer experience disconnected from the business behind it. | Operational systems | Integration | Digital experience | ||||
| Something that already works.Nothing new |
Sometimes the right intervention is no new technology: a step that changes owner, a tool already paid for used properly, a report nobody reads that stops. When that is the answer, we say so.
After it exists
The work continues once the system runs.
People are part of the system.
Adoption is designed, not announced. The people who use a system shape it, and we prefer to try a first version against real work before it replaces anything.
The business will keep changing.
New services, a second location, another person taking orders. We build so the next change fits where the work moved, without rebuilding what works.
Start with the business.
You don’t need to arrive with a technical specification. Tell us how the work runs today, where it gets stuck and what you have already tried. Deciding what technology belongs is our part.