Smoother processes
Reduce manual steps, bottlenecks and repetitive work in everyday workflows.
Local AI for SMEs and professional firms
Build an internal AI system that solves real problems, protects data and makes your team more independent, rather than selling yet another piece of software.
I start with your processes: documents, invoices, CRM, reports, marketing, sales, management accounting and repetitive activities. Then I establish whether you need cloud, private cloud or a local or on-premise solution with AI agents and internal applications.
From problem to solution
The market does not need another generic promise about artificial intelligence.
It needs tools that solve measurable problems: reading documents, reducing repetitive work, producing reports, checking data, supporting sales and administration, and building internal applications without constant dependence on external platforms.
My work is to turn these processes into AI agents, dashboards, automation and custom software. When speed matters, the cloud can be used. When privacy, sensitive data and control are decisive, we work with local or on-premise architectures.
Reduce manual steps, bottlenecks and repetitive work in everyday workflows.
Extraction, classification, verification and reporting from documents that currently take hours to process.
Turn archives, manuals and knowledge bases into answers the team can consult.
A system that performs tasks within defined rules and processes, rather than a generic chat.
Excellence and durability
A useful AI agent comes from a reliable structure: correct sources, clear rules, traceability, error handling, data protection and interfaces that make sense to the people using it every day. A well-written prompt alone is not enough.
Every business has different tools, documents, habits and constraints. That is why the work starts by analysing your actual workflows: invoices, emails, Drive, spreadsheets, reports, CRM, internal procedures and repetitive activities.
I start with a concrete use case, validate it with a prototype and only then select the most suitable architecture: cloud, private cloud or local. Hardware and theory follow the needs of the project.
AI operating system
A well-designed agent should save time every week, reduce errors and make business information easier to find, check and use. Its usefulness needs to last beyond the first five minutes.
Invoices, contracts, procedures, reports and archives become searchable and usable.
Small software tools, CRM, dashboards, workflows and specialist applications built around your process.
Training and support to make the business less dependent on external providers.
Where it makes sense, AI runs locally: data, permissions, logs and models stay under control.
The recommended approach
We map processes, documents, software, risks and opportunities. Output: 3–5 priority use cases and a concrete roadmap.
In 30 days, we build an agent or small application addressing a real problem to measure value, limitations and internal adoption.
Temporary support for local or on-premise architecture, team training, governance and progressive agent development.
First step
Tell me where you lose time: documents, data, reports, communication, internal processes, checks or repetitive activities. I will help turn an operational problem into a sustainable AI programme: an audit, a prototype and then independence.