Custom web applications

At some point, the process outgrows the spreadsheet. Two people edit the same file, nobody knows which version is the right one, and someone’s typo costs money. When that happens regularly, the problem is no longer a lack of discipline. It’s the lack of a system.

When it’s worth replacing the spreadsheet with a system

A spreadsheet is an excellent, inexpensive tool, and replacing it too early is a waste. The signs that the time has come are concrete:

• More than one person needs to edit at the same time, and you’ve already lost data to overwrites.
• You need to know who changed what and when, for an audit, for a contract or to find where an error came in.
• Different people should see different things, and today everyone sees everything.
• Someone spends hours every week copying data from one place to another.
• The spreadsheet has grown big enough to freeze or get corrupted.
• A client or supplier needs access to part of it, and today that’s handled by email.

If none of these applies to you, you don’t need a system. We’ll tell you that on the call, and we don’t charge to say it.

What we build

Internal dashboards

The screen your team works in all day. Records, approval workflows, owner assignment, change history. The quality bar here isn’t visual. It’s how many clicks the most repeated task takes.

Client and partner portals

Where people outside the company look up their own data without having to ask. Order status, documents, invoice copies, process tracking. This is usually the project with the fastest payback, because it cuts support volume.

Integrations

Getting your ERP, CRM, payment gateway, marketplaces and shipping carriers to talk to each other. It’s almost always the core of the project, and the real work is in error handling: what happens when the API on the other side is down, responds slowly or returns inconsistent data.

Automations

The task that runs on its own overnight: import a file, reconcile, generate a report, send an alert when a number goes out of range. It trades people time for machine time.

How it starts

It doesn’t start with screens. It starts with the process as it works today, exceptions included. That’s where the undocumented rules live, the ones that sink projects in the final stretch.

1. Mapping. We sit with the people who run the process today, spreadsheet open. We write down the normal path and every “but when X happens, we do it differently.”
2. Data model. What entities exist, how they relate, what’s required. This is the hardest decision to reverse later, so it comes before the interface.
3. Vertical slice. We build one complete end-to-end path, however small, and put it in the hands of the people who will use it. It tells you more than any clickable prototype.
4. Expansion driven by use. The next releases follow the order of what hurts most, not the order of the original list.

About migrating your existing data

This step is always underestimated. Spreadsheet data comes with CPF numbers (Brazil’s individual taxpayer ID) in three formats, dates in two, names with stray spaces, duplicates nobody knew existed, and a notes field holding critical information in free text.

We treat migration as its own phase, with a report of what came in, what was corrected and what needs a human decision. A new system with dirty data is an expensive way to lose your team’s trust in the first week.

What’s included by default

• Authentication with role-based permissions.
• Audit log: who did it, what changed, when.
• Automated backups with tested restores.
• A staging environment separate from production.
• Error monitoring that alerts us when something breaks, before a user reports it.
• Documentation of the architecture and the deploy process.
• Personal data handled in compliance with the LGPD (Brazil’s data protection law), including legal basis and retention periods.