Home / Solutions

Useful technology, designed around your constraints.

Sometimes a Sasly product fits as it is. Often the work is specific to one organization, and we build it. These are the four kinds of project we take on.

Private AI and knowledge systems

Make internal knowledge useful while keeping data boundaries clear.

Who it is for
Organizations with years of documents on shared drives, and staff who spend time looking for them.
What we build
Document ingestion, full-text and meaning-based search, results limited by each person’s permissions, and question answering from a local model where it is wanted.
How data is handled
Documents are indexed where they are. A local model is the default. Any use of a cloud model is a setting you turn on, for requests you define.
What we do not promise
That a model’s answers are always correct. Answers cite their sources so that a person can check them.
A typical first step
Index one department’s shared folder and let its staff search it for a few weeks.
Related product
Sasly Knowledge

Secure workflow automation

Sort, classify, summarize and review email and documents, with a person confirming.

Who it is for
Teams that receive a steady volume of email and documents and handle each one by hand.
What we build
Intake and sorting, classification, summaries, routing to the right person, and a review step before anything is sent or filed.
How data is handled
Processing is local first. Sensitive content can be redacted before any optional outside processing, and every automated step is logged.
What we do not promise
That classification is never wrong, or that sensitive messages are sent automatically. A person stays in the loop.
A typical first step
Automate the sorting of one shared mailbox and measure how often staff correct it.
Related product
Sasly Email

Application modernization

Modernize critical workflows without losing operational control.

Who it is for
Organizations that depend on an older internal application that few people can still maintain.
What we build
A web application and API to replace it, with role-based access, exports, data migration, and documentation so that it can be supported.
How data is handled
Existing data is migrated with a verification step, and the old system stays available until the new one has been accepted.
What we do not promise
A date before we have seen the existing system. We scope first.
A typical first step
A short review of the existing application, its data and the people who use it.
Related work
Legacy application replacement for a municipality

Data governance and security

Information boundaries, permissions, audit and retention, designed in from the start.

Who it is for
Organizations that must be able to show who can see which information, and who did.
What we build
Permission models, audit logging, retention rules, secrets handling and backup, as part of the systems we deliver.
How we work
We write down where data lives, who can reach it and what leaves, and we review that with you before development starts.
What we do not promise
Certifications we do not hold. We are not a managed security provider, and we do not offer penetration testing as a service.
A typical first step
A data map for one system: what it stores, who uses it, and where copies go.

Not sure which of these your problem is?

Most are a mix. Describe the situation and we will tell you how we would approach it, or that we are not the right fit.