aprivas is a specialised team
Our work happens in a small, experienced team with focus areas in automation, platform engineering, security and AI. We work systemically, document carefully and hand over reliably. Started in 2015 as a sole proprietorship, we have operated as aprivas UG (haftungsbeschränkt) since 2016.
Who is behind aprivas
Founding Architect
Markus
Connects classic Linux infrastructure with sovereign-AI and multi-agent architectures.
Office & communications digitalisation
Sabine
Digitalisation of internal office and team-communication processes: document workflows, consistent tool landscapes, reliable handovers into existing structures.
Project development & planning
Stefan
Konzeption und Phasen-Planung komplexer Vorhaben, von der Anforderung über die Entscheidungs-Architektur bis zur belastbaren Übergabe.
Algorithmic systems & infrastructure automation
Max
Algorithmische Trading- und Entscheidungs-Systeme sowie reproduzierbare Infrastruktur-Automatisierung auf Plattform-Ebene.
Our AI agents
Alongside our four staff members we use our own AI agents. They are not persons and not employees, but software from our own agent system. They support research, triage and operations. Responsibility and decisions rest exclusively with the named persons.
Ida
AI agentAgent: inbox triage & correspondence drafting
Sorts incoming messages and prepares draft replies, but sends nothing externally itself: every reply is reviewed and sent by a human. It makes no decisions.
Scout
AI agentAgent: research & source verification
Collects and verifies sources for project and market questions, every statement with provenance.
Meno
AI agentAgent: coordination & distribution
Accepts tasks, breaks them down and distributes them to the right specialist agents, with a return path through the review chain.
Opix
AI agentAgent: operations monitoring & execution
Watches systems and runs, executes approved routine tasks and reports deviations to on-call.
Palo
AI agentAgent: document filing & order
Sorts, tags and archives documents by fixed rules, findable even years later.
Weitere Agenten
AI agentsAgent group: scheduling, document flows, knowledge upkeep, internal routines
Behind the named ones works a group of further specialist agents. Each with a narrow mandate, each logged, none deciding alone.
Prisma
AI agentAgent: quality & rule checking
Checks the work of the other agents against rules and checklists before a human sees it.
Which competencies the team covers
- 01Platform & automationCI/CD, infrastructure-as-code and operating platforms, covered by the founding architect and project planning.
- 02Security & governanceBaseline protection, audit preparation and evidence, anchored across the whole team.
- 03AI systemsLocal AI architectures, RAG and agent systems, led by the founding architect.
- 04Digitalisation & DMSDocument processes and office digitalisation, covered by office digitalisation and planning.
Wie wir arbeiten
- 01We write down what we do: documentation is part of the result.
- 02No project without a handover plan: systems must run without us.
- 03Honest estimates, even when they are inconvenient.
- 04We use standard technology consistently before building something new.