How we work

What you send us

An email to office@vsoft-design.com or a call to +40 752 287 722, with:

• the work process concerned, in brief: what is done, who does it, which documents and data pass through it;
• the sector, for example insurance, lending or financial services;
• the approximate number of participants, if it concerns laboratories or a training programme.

There is no need to send real data or personal data: a description of the process is enough for a first discussion.

What you receive: a written offer

Based on the description, Virtual Soft sends a written offer. The offer sets out:

• the scope of the work: the process covered, the laboratories or the training programme, and what stays outside them;
• what is delivered at each stage;
• the schedule of the stages;
• the price.

Everything that depends on the project — the number of theory and laboratory hours, group sizes, programme length, the support period and its terms — is set in the offer and the contract, not on this site.

Virtual Soft S.R.L. is not registered for VAT: the price in the offer contains no VAT, and the invoice adds none.

What each stage delivers

The stages are those of the approach, in this order.

1. Analysis. A written description of the process, its constraint and the success indicators, worked out with the people who run the process.
2. Design. The laboratory scenarios, the synthetic datasets, the provoked failure cases and the assessment criteria. The client validates the scenarios before development. The demo Laboratory with a provoked failure shows how such an exercise is built.
3. Development and pilot. The application, built as a shared engine plus a domain pack, on mature open-source components where they exist. A pilot group works through the laboratories, and its observations feed back into design.
4. Delivery and transfer. The source code, a non-exclusive, perpetual, transferable licence, the technical and user documentation and a knowledge-transfer session. For training programmes: the course materials, laboratory sheets and assessment instruments. Examples of course materials are on the Training materials page.
5. Support. Fixes, content updates and assistance for the team that took over the system, under the terms set in the contract.

Where it runs

Training programmes run online, for groups. The delivered application stays with the client: with the source code and the licence, the client's team can run, change and extend it, and can continue development with anyone. Where the application runs is agreed with the client.

When a project includes AI components, the choice is between open models hosted locally and EU API services, judged on cost, confidentiality and control. The demo Cost per task for model configurations shows the cost side of that choice.

Data and the GDPR

• Laboratories work on synthetic data generated by the application; it contains no real records.
• When it works with the client's data, Virtual Soft acts as a processor (GDPR Art. 28), under a written agreement, and processes the data only on the client's instructions. Learners' data in the training platform is personal data and is processed under the same terms.
• The applications do not grade learners automatically with AI models and do not use biometric proctoring.
• The demos on this site run in your browser and send no data to the server; the details are in the privacy notice.

What we do not offer

What Virtual Soft does not offer — authorisations and certifications it does not hold, automated grading of learners — is listed on the Services page, in the paragraph "What we do not offer".

Frequently asked questions

What is a digital laboratory on synthetic data?

It is an exercise built on a real process of a sector, in which learners work with generated data rather than real records, and meet failure cases provoked on purpose. Their results can be checked and go into the learner's portfolio. The demo Laboratory with a provoked failure shows such an exercise.

Is real data used?

Not in the laboratories. The data is synthetic and generated from a seed, so the same seed produces the same dataset. Synthetic data reproduces only the relationships that were modelled and does not replace validation on real, authorised data. The Synthetic data generator with validation shows how such a dataset is generated and checked.

What does the client receive at the end?

The source code, a non-exclusive, perpetual, transferable licence, the technical and user documentation and knowledge transfer to its team. For training programmes it also receives the course materials, laboratory sheets and assessment instruments.

How is data handled (GDPR Art. 28)?

Virtual Soft acts as a processor, under a written agreement, and processes data only on the client's instructions. The laboratories use synthetic data, which contains no personal data.

Is AI used to grade learners?

No. The applications do not grade learners automatically with AI models and do not use biometric proctoring.

Where do the demos run?

In your browser. The generated data and the values you enter are not sent to the server, and the same seed always gives the same result. All demos are on the Demos page.

Can I check the company?

Yes. Virtual Soft S.R.L. is registered with the Romanian Trade Register (ONRC) under number J2006000901248 (EUID ROONRC.J2006000901248) and has the tax ID (CUI) 18736264. The details can be checked publicly with ONRC and ANAF; the full list is on the Company page.