Software development

We work in phases, with deliveries that add value from the beginning, and with a fixed idea: that what we build can be maintained and evolve for years, by us or by whoever comes next.

What problems do we solve

• You need more than a corporate website. E-commerce, private portals, customer areas, or intranets connected to your internal systems. This is what we do in digital platforms and e-commerce, with cases like the evolution of Bodeboca's e-commerce.
• Your systems do not communicate with each other. Manual exports, duplicated data, failing synchronizations… We design APIs and integrations with ERP, CRM, payment gateways, and third-party services so that data flows reliably and traceably. More detail in system integration and software architecture.
• You want your product to incorporate artificial intelligence. Semantic search, content classification, assistants connected to your documentation, or process automation, integrated as a part of the system and not as an ornament. This is what we call software with applied AI.
• You have an ongoing project that needs a handover or reinforcement. We can take on the complete development or integrate with your internal team, providing architecture, code, documentation, and technical expertise.

How we work

1. Objectives and scope. We start from your business objectives and operational needs to define the functional scope and architecture. If there are pending fundamental decisions, we resolve them first with a phase of technology consulting.
2. Deliveries in phases. We prioritize what adds the most value and deliver it incrementally, so you can validate it soon and adjust the course with real information.
3. Continuous quality. Testing, code review, and attention to what happens in production in each iteration, not just at the end.
4. Documentation and transfer. The code and decisions are documented. We work with open technologies and standards so that you are not dependent on us.

How we use AI agents in development

We use AI agents and assistants as part of our engineering methodology, because when used properly, they allow us to progress faster in tasks such as exploring existing code, writing tests, or preparing repetitive changes. But speed does not replace judgment, which is why we apply clear rules:

• All code is reviewed by a person. No change generated with the help of AI reaches your project without human review, just like any other code.
• Your data does not leave without your permission. We do not send your clients' data or confidential information to AI tools without your express authorization.
• You decide which tools are used. If your organization needs it, we limit the work to the AI tools or models that you approve.
• The same quality requirements. AI-assisted code undergoes the same tests, the same review, and the same standards as the rest.
• Our responsibility. The tool does not sign the work. We are responsible for the outcome just as always.

Do not confuse this with software that incorporates AI: here we are talking about how we build, not what the product does.

Related case: implementing Checkout.com in a legacy e-commerce

We have implemented Checkout.com in a client's e-commerce, a platform with over ten years of evolution, a lot of legacy code, and little documentation. It operates in five markets (three active and two in preparation), with mobile applications and recurring payments. The new payment methods, such as Bizum, Apple Pay, or Google Pay, have had to coexist with the previous ones. The launch has been done without service interruptions and with a single minor incident, resolved without significant impact on users.

Frequently asked questions

Will I depend on you to maintain what you build?

No. We work with open technologies and standards, and we document the code and decisions so that you can maintain it with your team or whoever you choose.

What technologies do you work with?

Mainly with Drupal, PHP/Symfony, and Python, as well as decoupled architectures and integrations with third-party systems. We choose the technology based on your context, not the other way around.

Do you write code with artificial intelligence?

We use AI agents as support, always with human review of all code and without sending your data without permission. If you need it, we limit ourselves to the tools or models that you approve. The responsibility for the outcome remains ours.

Can you continue a project that another provider has started?

Yes. Many of our projects start from existing platforms. First, we analyze the state of the code and then propose how to proceed with the least risk possible.

Can you work alongside our internal team?

Yes. We can take on specific parts of the development, provide architecture, or reinforce the team, adapting to your way of working.