Does NoNetSearch use AI?
It uses modern retrieval technology to understand meaning rather than relying only on exact keywords. There is no generative AI model answering questions or summarising documents in the search path. It is search, not chat.
Where does our data go?
NoNetSearch runs on infrastructure you control. Your document corpus is not indexed, stored or processed on NoNetSearch-hosted servers.
How is it different from generative AI search tools?
NoNetSearch retrieves and ranks genuine source passages. It does not generate an answer on the user’s behalf. This gives organisations an option when the desired outcome is search without introducing a generative system at query time.
Do we need a GPU?
Not for Community or Plus. OCR of scanned PDFs is CPU-based and does not require a GPU. Enterprise uses a GPU only for image and visual-content search.
Can it run on Windows Server?
Yes. The Linux-based Docker image can run on Windows Server through Docker Desktop with WSL2. Ubuntu is also supported.
How is NoNetSearch installed?
We provide the Docker image and offline licence. Your technical administrator or MSP deploys it on your server, connects selected network sources and explicitly approves what enters the index.
How long does installation take?
A technical administrator already comfortable with Docker can typically complete the initial setup in under a day. The first scan duration depends on document volume, formats and available hardware.
How does the trial work?
The 15-day full-function trial begins when the first document is ingested, rather than on the installation date.
Does the licence require an internet connection?
No. A signed licence certificate is verified locally, with no phone-home requirement or ongoing internet dependency to keep the licensed deployment operating.
How many people can use it?
Licensing is based on capability tier rather than seat count. Practical concurrency depends on the server specification and will be considered during deployment sizing.
How quickly are results returned?
Response time depends on hardware, index size and concurrent usage. We prefer to size and benchmark against your expected environment rather than publish a universal speed claim.
What happens if we cancel?
Your original documents remain on your hardware. NoNetSearch has not hosted a copy of the corpus, so there is no vendor-held document store to return or delete.
Does NoNetSearch replace our existing file permissions?
No. It works alongside them. Search shows the relevant source and location, while opening the underlying file uses your existing access route and permissions.
Are new documents indexed automatically?
Newly discovered folders and files surface for administrative review before inclusion. This prevents content from being silently swept into the searchable index.
