Your files answer, and every answer shows its source
An assistant that reads your company's policies, contracts and manuals, answers your staff from them, and shows where each answer came from. If it can't find the answer, it says so rather than guessing.
Last updated:
The problem
The information is in a file somewhere, but nobody knows which one. So people ask a colleague, or answer from memory.
How we work
- A demo on your files: 3 to 5 files, in a test environment, deleted afterwards.
- A pilot in one department: with around 100 test questions and a success criterion agreed in writing before we start.
- Go-live: an hour of training, a user guide, and a panel listing the questions it couldn't answer, so the missing files can be added.
- Monthly report: what was asked, what was answered, and what needs a new file.
Fact sheet
- Timeline
- The demo comes before any commitment. Pilot and rollout timelines are set in the written proposal
- Deliverables
- The assistant with chat and search, an admin panel for files, a guide, a monthly report
- Includes
- QMM DocAI (ask your files) · an assistant inside every system · chatbots · book digitisation and text clean-up · scholarly and heritage works
- Markets
- Saudi Arabia and Egypt
- Languages
- Arabic and English
Digitising books and scholarly works
We start with one chapter as a sample, so you can review the result before the rest. Religious content is reviewed by a qualified scholar and linguist, not by software alone.
Questions
Does it answer from outside our files?
No. It answers only from your files, and says so when it can't find the answer.
Does it give the page number?
It shows the file and the passage it relied on. Page numbers depend on how each file was prepared, so we don't promise them.
What happens to the demo files?
They stay in a test environment and are deleted after the demo.