Automation & Responsible AI
SERVICES

We help you lighten your business processes: less data entry, less retyping, fewer oversights. When it brings real value to your work, we integrate artificial intelligence (AI) in a responsible, well-framed way. The software we use is open source, adapted to how you work and stays under your control.

Illustration of connected tools and automated tasks

What we can automate

A few examples. We start from your tools and decide with you, case by case, what is worth automating.

Your repetitive tasks

Recording a contact, tracking your services and tasks, filing incoming invoices, scheduling newsletters and posts…

Your documents

An assistant finds information in your procedures, contracts or past files, and points to the passages (RAG).1

Your tools, connected

n8n connects natively to Odoo, HubSpot, Brevo, Google Workspace or Microsoft 365, and to any tool with an API.

Example: tracking calls for projects and grants

Every week, this flow built with n8n (an open-source automation tool) gathers the calls for projects published by funding bodies. A keyword filter removes most of them, then a small AI model, given a description of your nonprofit, sorts the rest. The leads it keeps go into your tracking tool, with a weekly summary.

n8n workflow: every Monday, three sources of calls for projects are read (Wallonia, Wallonia-Brussels Federation, European Union), merged, de-duplicated, filtered by keywords and sorted by a small Mistral model according to the nonprofit's profile; relevant calls are added to tracking (NocoDB), grouped, and sent as a weekly summary.

Data, security, sobriety

In Europe, or on your premises

Your tools and data hosted in Europe, on a rented server (VPS) or in your offices, including for AI (Mistral, Ollama).

Security from the start

Role-based access rights, human approval of sensitive actions2,3 and regular backups, in line with the GDPR.

Sized to the need

Open-source tools proportionate to the use, a simple rule when it is enough and, for AI, the most compact model suited to the task.4

How we work together

A structured, five-step approach, carried out with your teams.

  1. 1. Current situation

    Your processes, tools and constraints, described together in plain language.

  2. 2. Audit and priorities

    What is worth automating, what is not, and where to start: a plan decided with you.

  3. 3. Implementation

    Built on your real cases and tested alongside your current way of working before going live.

  4. 4. Support

    Your teams get to grips with the tool. If AI is involved, a workshop and a usage policy set the framework.5,6

  5. 5. Monitoring and maintenance

    Hosting, updates, alerts on errors7 and changes as your needs evolve.

Let’s talk about your processes

Tell us which tasks take up your time and which tools you already use. In a first conversation, we look together at whether automation makes sense, and how to put it in place.

Write to us
Sources 7
  1. Chen et al., 2024, “Benchmarking Large Language Models in Retrieval-Augmented Generation”: knowing when to say “I don’t know”.
  2. EU AI Act, Article 14: human oversight.
  3. GDPR, Article 22: no fully automated decision that significantly affects a person.
  4. Luccioni et al., 2023, “Power Hungry Processing”: energy use varies widely with the task and the size of the model.
  5. Microsoft and LinkedIn, Work Trend Index 2024: 78% of people who use AI at work bring their own tools.
  6. EU AI Act, Article 50: informing people that they are interacting with an AI system.
  7. n8n documentation: error handling and execution logs.