FAQ

Frequently asked questions.

Straight answers about our data and AI work: what we build, how a project runs, and how we handle your data. Not here? Ask us.

About Growing Pai

What does Growing Pai do?

We are a data and AI consultancy in Ghent, Belgium. We turn your data into personal AI that does real work: reading orders, onboarding clients, answering mail and predicting what is next.

What does Pai stand for?

Personal AI: AI shaped around your data, your tone and your way of working. Growing stands for continuous growth, and for growing the pie rather than fighting over slices.

Who do you work with?

Organisations in manufacturing, HR and payroll, the public sector, energy, media, logistics, automotive and food. Examples are Attentia, Versele-Laga, the City of Ghent, the European Environment Agency and BD Media.

See our cases
Which technology and partners do you work with?

We are a Databricks and Microsoft partner. We build on Azure, Databricks, Microsoft Fabric, Power BI, Azure AI Foundry and OpenAI, and LangGraph for agents.

Data platforms, integration and dashboards

What is a data lakehouse, and do we need one?

A lakehouse brings your data together in one place, with one version of the truth for reporting and AI. We build them on Azure, Databricks and Microsoft Fabric, for SMEs and enterprises. A Microsoft Fabric lakehouse for Empwr Nutrition was up in two months.

Read the Empwr Nutrition case
Databricks, Microsoft Fabric or Power BI: which should we choose?

It depends on your existing landscape, data volumes and your team. We compare the options openly, including the ones that do not involve us building anything, so the decision is yours and an informed one.

Can you connect SAP, our CRM and other systems?

Yes. We bring SAP, CRMs, IoT data, files and APIs together, clean them and make them ready for analytics and AI. For Empwr Nutrition we staged SAP data with Azure Data Factory.

Can you build dashboards my team will actually use?

Yes: Power BI, Fabric or Databricks dashboards, or custom web dashboards inside your own portal. We aim for self-service, so the next question does not need a ticket. At BD Media we set up a network of internal Data Champions.

Read the BD Media case
What is the difference between batch and streaming data pipelines?

Batch pipelines move data on a schedule, for example every night. Streaming pipelines carry it all day long. For an automotive group in Ghent, one config-driven framework on Databricks handles both.

Read the case

AI agents, automation and document intelligence

What is an AI agent?

A digital colleague that reads, checks, decides and acts across your systems, and asks a human when it is unsure. Business rules stay separate from the AI, a person approves what matters and every step is logged.

Can AI automate order intake from PDF, Excel and email?

Yes. At Versele-Laga, about 80% of order emails go from PDF, Excel and scans into the ERP automatically, in every layout and language, and more than 5,000 orders have been processed. People only review the orders the system is unsure about, and no customer had to switch to EDI.

Read the Versele-Laga case
What is document intelligence?

Software that extracts, classifies and matches information from messy documents, in any layout or language, and hands it to your systems.

Can AI speed up client or employee onboarding?

Yes. The Attentia Welcome Agent turns messy client Excel files into validated payroll data and cut onboarding time by 35%, with a human approving what matters and a full audit log.

Read the Attentia case
What can machine learning and predictive analytics do for us?

Forecasts that show what is coming. For the City of Ghent we built a model on 20 years of road data that predicts road quality three years ahead across 1,403 km of roads.

Read the City of Ghent case
Do we need deep learning or computer vision?

Not always. For the European Environment Agency we compared a deep-learning model on satellite images with classic machine learning to predict air quality. The best technology is the one that fits, and sometimes that is a well-tuned classic model at a fraction of the cost.

Read the air quality case

Private chatbots, privacy and trust

Can we have a private ChatGPT on our own documents?

Yes. We use retrieval-augmented generation (RAG): employees ask questions in plain language and get answers from your internal documents, with sources. Access rights are respected, and when there is no answer the assistant says so.

Read the private ChatGPT case
Is our data safe, and where does it go?

We design for GDPR, data location, access rights and audit trails from the start. At Attentia the data stays in the client's own Azure tenant, rules are kept separate from the AI and every step is logged.

How do you deal with GDPR and the EU AI Act?

We check which rules apply early, because it is far cheaper to design for them than to retrofit. That covers GDPR, data location, access rights, audit trails and the EU rules on AI.

How do you keep AI answers reliable?

Answers cite their sources or say there is no answer. Business rules sit outside the AI, a human reviews what the system is unsure about, and we measure and report results.

Will AI replace our people?

No. Humans guide the AI. We take the repetitive work off your plate so your people can focus on the interesting parts.

Working with us

How does a project work?

In four steps: Pick one use case worth doing first, Prove it with a fast first version, Ship what works with the right guardrails, and Improve it by measuring impact and listening to users. We start small and show something real quickly.

Read our approach
We do not know where AI fits. Where do we start?

With an AI opportunity workshop: a hands-on session with your team to find where AI pays off first, and a realistic first step. The City of Ghent started with a workshop and three ideas, then piloted the best one.

Book a workshop
How fast can we see something working?

We build with AI coding agents and automated tests and reviews, so a working first version arrives in days, not quarters, with engineers accountable for every line.

Should we buy, build or blend?

Off the shelf where it is good enough, custom where it is your edge, and often a mix. We lay out the options and what each costs to live with before anything is built.

Do you support us after go-live?

Yes. Run and grow covers ongoing support, troubleshooting and new use cases, so your system keeps up with your business.

Do you train our teams on AI?

Yes. We offer training, playbooks and policies so your people use AI daily and safely: what data may go where, who approves what and how results are measured. We also gave a client a five-day Databricks training.

How do I get in touch?

Email axel@growingpai.com, call one of our co-founders or business developer, or book a call.

Contact us

Still have a question?

Tell us what slows your team down. We will tell you honestly whether AI can help.