Cave Canem

Custom web platforms shaped around how your organization actually works: content management, e-learning and e-commerce.

We build web platforms from the data model up, not by stacking plugins until something breaks. That means content management systems editors are happy to use, e-learning platforms with real course logic, progress tracking and assessment, and e-commerce that copes with large catalogs, complex pricing and the systems behind the shop.

Desktop software for macOS, Windows and Linux, from point-of-sale systems to tools wired into your existing infrastructure.

For work where a browser tab is not enough: point-of-sale systems, back-office tools and operator consoles. We ship on macOS, Windows and Linux, and connect the application to what is already there, including peripherals, databases, and third-party or in-house systems.

Linux and Windows servers installed, configured and documented, so they keep running after the project ends.

We deploy and configure Linux and Windows servers, both for the software we build and for the software you already run. Deployments are repeatable, configuration is written down, and nothing depends on one person remembering how it was done.

Firmware for ESP and other microcontroller platforms. We write the software; the hardware design stays with you.

We write firmware for ESP boards and other microcontrollers: reading sensors, handling connectivity, running device logic and updating it in the field. We do software only. The circuit and the board come from you or your electronics partner, and we make the code fit them.

Virtual and augmented reality applications built for a job, such as training, visualization or guided procedures.

We build VR and AR applications where immersion does real work: training, product and space visualization, guided procedures and interactive presentations. The work covers interaction design, 3D content integration and deployment to the headsets or devices your users actually have.

Connecting closed, legacy and undocumented systems to the rest of your stack, and getting your data back out of them.

When a system has no documentation, no API or no vendor left to call, we work out how it behaves and connect it to everything else. Typical jobs include decoding undocumented protocols and file formats, recovering data from closed or end-of-life software, and building bridges between systems that were never meant to meet. We work only on systems you own or are authorized to access, for interoperability and continuity.

LLM pipelines with a defined job inside your systems: document search, data extraction, classification and automation.

We build AI pipelines that do one clear job inside the systems you already have: answering questions from your documents, turning unstructured input into structured data, classifying and routing requests, drafting text for a human to approve. Each pipeline is tested on your real data before it goes live, and designed so the model can be swapped when a better one comes along.

A straight technical opinion before you commit: architecture, technology choices, audits of existing systems and vendor proposals.

Sometimes the useful deliverable is a decision, not code. We review existing systems, plan architectures, assess vendor proposals and help you decide what to build, what to buy and what to leave alone. You get a clear written recommendation, even when the answer is to do nothing.