02.07AI

Knowledge assistants

Knowledge assistants from Hot Rocket Software, a team of software, website and digital marketing specialists based in the UK, with 19 years in IT. A private assistant that answers questions from your own documents, policies and records, with citations back to the source, and respects who can see what.

INPUTMODELREVIEWHUMAN IN THE LOOP02.07 KNOWLEDGE ASSISTANTS

02.07.0

What is knowledge assistants?

A knowledge assistant answers questions from your own information, such as policies, procedures, contracts, product manuals and past tickets, and shows the source for every answer. It respects existing permissions, so people only get answers from documents they are allowed to see. It is the fastest way to stop the same questions landing on the same people. It is one of the practical AI tools we build as part of our everyday work, set up and supported by real people on our team.

See where knowledge assistants fits in our work →

02.07.1

Who is it for?

  • Teams where knowledge lives in a few people's heads and a shared drive
  • Support teams answering the same questions repeatedly
  • Regulated businesses that need answers traceable to a source

02.07.2

From a small business to a large organisation

The same team takes on a single tool for a small business and long-running work for large companies and the public sector. Here is what each end looks like.

For a small business

An assistant that answers staff questions from your handbook and price list, and says when it does not know.

For large companies and the public sector

Search and question answering across large document estates, respecting existing permissions and showing the source of every answer.

  • Policy and procedure assistants
  • Permission-aware search
  • Large document migrations
  • Answer quality monitoring

How we work with enterprise and the public sector →

02.07.3

What you get

  1. 01Connection to your sources (SharePoint, Google Drive, wikis, ticket history)
  2. 02Retrieval with citations on every answer
  3. 03Permission-aware search
  4. 04An admin view of questions asked and gaps found
  5. 05Hosted in the UK or on your own infrastructure if required

02.07.4

How we approach it

  1. Discover

    Collect the questions people actually ask.

  2. Scope

    Which sources, which users, which permissions.

  3. Build

    Ingest, index, answer with citations, evaluate.

  4. Launch

    Pilot with one team, measure answer quality.

  5. Support

    Fill gaps the questions reveal; keep sources fresh.

02.07.5

Typical stack

  • TypeScript
  • Python
  • PostgreSQL + pgvector
  • LLM APIs
  • MCP

02.07.6

Questions

Is this just ChatGPT on our files?

No. It retrieves the relevant passages from your sources, answers only from them, cites them, and respects document permissions. General chat tools do not do all of that by default.

Where is our data stored?

Where you need it to be. We can host in UK data centres or on your own infrastructure, and use model providers that do not train on your data.

What if the answer is not in our documents?

It says so, rather than guessing. Those gaps are logged so you can fill them.

02.07.7

What could this look like for you?

Example ideas, not client case studies.

More ideas we can build →

T+99Next step

Talk it through with us

A free 30-minute first chat, with no obligation. Tell us what is getting in the way; we tell you honestly what we would do, roughly what it would take, and whether it is worth doing at all.