Skip to content
TREN
Start a project

Company knowledge assistant

Ask questions to your own documents and see the page each answer comes from.

Typical targets; the real figure is measured in the pilot.

  • Within secondstarget: time to find an answer across hundreds of pages
  • Source pagenext to every answer, so you can check it
  • 24/7target: the team can ask at any hour

The problem

Knowledge is scattered across folders, emails and a few people's heads. A new hire asks three people to find a procedure. Looking for one clause in a contract or manual takes half an hour.

Who it is for

  • Manufacturing and technical service firms
  • Law and accounting offices
  • Construction and engineering firms
  • Multi-branch and multi-team businesses
  • Training and HR teams

What it does

  1. Answers from your documents

    Reads manuals, contracts, procedures, PDFs, Word and spreadsheet files. Answers come from your documents, not from general knowledge.

  2. Shows the source

    Each answer says which page of which document it came from. One click to check.

  3. Access by department

    Not everyone sees HR documents. You decide who can reach which documents.

  4. Reachable where you work

    Ask from a web page, Slack or Microsoft Teams. The team stays where it already is.

  5. Documents stay current

    When a file is added or updated, the assistant picks it up. Documents are kept current so it does not answer from an old version.

Savings estimate

What routine work costs

Set three numbers. If 60% of the routine work is automated, see roughly how much time and money is left each year.

Department

Example routine task: Invoice entry, reconciliation, collections follow-up

3
3
₺45,000

Automation rate is fixed at 60%, with 22 workdays a month.

Estimated gain

Yearly saving
₺364,500
Hours saved per year
1,426 hours
Full-time equivalent
0.7

This is an estimate; the real figure is measured in the free analysis.

Ask for a process analysis

How we deliver it

  1. 01

    We learn your documents

    Which documents exist, where they live, who should reach what. We settle it together.

  2. 02

    We prepare the documents

    We make files readable, organise them for search and set the access rules.

  3. 03

    We build the assistant

    We open access through web, Slack or Teams and test with the team's real questions.

  4. 04

    Roll out and improve

    We look at unanswered questions and add missing documents or settings.

Works with

  • Google Drive
  • Notion
  • Slack
  • Microsoft Teams
  • SharePoint
  • OpenAI
  • Claude
  • Cloudflare

Questions

What if the assistant answers wrong?

Each answer comes with its source page, so you or the team can check it. If the documents do not contain the answer, it is set up to say it could not find it instead of making one up. For contracts and legal topics we can require a person to approve the answer.

Where do my documents go, and is it KVKK-aware?

Under the business (API) terms of the model provider we use, data is not used to train models; we check the current terms with you before setup. You choose which documents are processed; we can leave out or mask those with personal data. We write down where data is kept during setup.

How long does setup take?

It depends on the number and order of the documents. A first working version, with a selected group of documents, is usually ready in 1-3 weeks. Other documents are added after.

What drives the cost?

The volume and order of documents, the number of users, the channels (web, Slack, Teams) and how complex the access rules are. We listen to what you need and give a clear quote.

More from this service

AI

Company knowledge assistant

Ask for a free analysis