A single virtual microscopy platform
Slides analysed one at a time, and images too large for the tools already in use.
Artificial intelligence, in our work and in your software
We use it on every project and bring it into the software we build. Neurally, an AI company with two products of its own, was also born from dotenv.
How does dotenv use artificial intelligence?
In two ways. In our work, on every project: it reads documentation and requirements and works alongside the developers while they write the code. And inside the software we build, to recognise images, transcribe and translate audio, classify text and content. Client data only passes through private AI services of our own.
AI in our work
We use it on every project, and we govern the tool: project choices stay with the team, and every release has a person who answers for it.
Requirements and documentation
AI reads technical documentation, requirements and meeting transcripts, and puts them in order for the team.
Development
The developers work with AI assistants while they write the code. dotenv answers for the code we release.
Client data
It only passes through private AI services of our own, never through an external AI service.
AI in the software we build
It helps where today a person repeats the same check many times a day: reading every listing before it goes live, writing up the minutes of a meeting by hand, looking for a detail in very high resolution images. The software makes the first pass, and the person looks at what the software flags.
Recognising images
Computer vision models that analyse images, very high resolution ones included, and flag what a person would otherwise have to look for by eye. The flag goes to whoever has to assess it.
Transcribing and taking minutes
Speech recognition that transcribes meetings and talks, live or afterwards, filters out background noise and sends the text to the participants. It can run on the client’s servers, so the audio does not leave.
Live translation
Translation into several languages during meetings, webinars and talks.
Classifying
Automatic classification of text, images and video: the software decides the clear cases, blocks the ones that break the rules and passes the doubtful ones to a person, with the analysis of what it found.
Inside the systems you have
The model is one part of the software: the results end up in the management system, the CRM or the platform you already use, not in a separate application.
Code written with AI
The questions worth asking a software house that uses AI to write code apply to us too: which security checks the code goes through before release, who reviews it and what they look for, how user input is handled, in which parts of the project AI assistants are used.
On copyright, the Italian law on artificial intelligence, passed in 2025, says that code written with the help of AI can be protected, if it is the result of a person’s intellectual work. The code we write for you is yours, with the documentation and the environments.
The rules on AI in products
Software with AI inside has to comply with the AI Act, the European regulation on artificial intelligence. The transparency rules apply from 2 August 2026: whoever talks to a chatbot has to be informed that they are talking to an AI system, unless it is obvious, and in software that is an interface choice. The rules on high-risk systems, after the AI Omnibus, apply from 2 December 2027, and from 2 August 2028 for AI inside regulated products such as machinery, toys and lifts.
In Italy, since 30 September 2026, Legislative Decree 160 of 2026 is also in force, aligning Italian rules with the AI Act on civil and criminal liability. That is why, when we build software with AI, we decide with you where the model and the data run, what the system records when it decides or suggests something, and what we hand over at the end of the project.
Frequently asked questions
Does my company’s data end up in an external AI service?
No. On client data we only use private AI services of our own.
Do you use AI to write code?
Yes, on every project, as support for the developers. dotenv answers for the code we release.
Where does the model run?
It depends on the data: on private AI services of our own, or on the client’s servers when the data must not leave.
How does it start?
With a call, free of charge. Then a visit to your site, to see how you work today and where AI can help.
Some projects with AI
What we have written about AI
- Software development with AI: does it cost less and go faster? What the studies say
- Is AI-generated code secure? What to ask the software house
- Code written with AI: who owns it? Copyright after Law 132/2025
- AI Act deadlines after the AI Omnibus: what already applies from 2026 and what is postponed
- D.Lgs. 160/2026 on artificial intelligence: what changes in civil and criminal liability
- SIMEST, 200 million for artificial intelligence in SMEs: requirements, grant and application
- Automatic learning: an introduction to Machine Learning
We bring AI into your software.