Your work, read properly
and pointed back at you.
askFinz reads the web itself rather than buying somebody else’s search results. That makes a partnership straightforward: what you publish is read as written, kept current, and quoted with your page attached. And where the material is yours alone, Indexer OS reads it on hardware you already own.
A person reads every enquiry, and replies come from partners@askfinz.com — usually within two working days. If you would rather see the thing working first, the sites being read right now are on the page, live.
Three ways this usually starts.
A partnership here usually starts in one of three ways. If none of them is quite you, say so on the form — the interesting ones rarely fit a category.
Publishers, universities, standards bodies
You publish something worth reading
Documentation, research, reference, reporting, filings. Your pages are read as written and answered from directly, with the passage carrying your page and your name back to the reader.
- Read as a document, not scraped as a page
- Every answer carries the address it came from
- You decide how it is read, and can stop it
Organisations with their own material
You hold knowledge nobody else has
Corporate pages, product documentation, public-facing sites and the knowledge clusters your teams have built. Indexer OS runs on hardware you already own and reads them where they sit.
- Runs on spare hardware — boot it and walk away
- Reads your corporate and public-facing pages
- Knowledge clusters kept current, not re-crawled
Platforms and technology partners
You want to build on the index
The index is ours rather than rented, which is what makes it something to build on: workspaces, agents and products can read from it directly instead of paying per web call.
- One index behind search, research, chat and the rest
- Records keep their type, not flattened into pages
- Talk to us about what you need it to do
When not to write
You just want a site read. You do not need a partnership for that. Put the domain in the queue and nobody has to be emailed.
You want to buy a data feed. askFinz does not sell index access as a standalone feed or API, and there is no marketplace to build on. That is a deliberate choice rather than a roadmap item.
You want your ranking changed. Being read is not the same as being promoted, and there is nothing to buy that moves you up an answer.
Specific things, not a relationship.
Read when you publish
Pages become answerable as they are read rather than after a rebuild, so a correction you make today is what a reader gets today.
Credited, every time
Answers arrive as a passage with its source attached. Your work is quoted and pointed at, not absorbed and paraphrased away.
Filed as what it is
A judgment, a standard, a filing and a paper are different kinds of document. They keep their identifiers, jurisdiction and dates instead of being flattened into web pages.
Indexer OS on your hardware
An appliance for machines you already have. It sets itself up, recovers itself and updates itself, so nobody on your side has to look after it.
Read what is published, only
Public pages are read as any visitor would see them. Anything beyond that happens only where you have asked for it and supplied your own access.
Documents read all the way through
Reports, filings and slide decks are converted to readable text and actually read, rather than linked to on the strength of a filename.
Visible on the live page
Your domain appears on the live status page once it is through, so what we say is happening is something you can check rather than take on trust.
It comes back to you
This is the part that is easy to miss. What you contribute does not leave and become somebody else’s product — the same shelf is read by the workspaces your own people work in, so a standard your team published is answerable in Research, and a filing you keep current is there when Finance builds a view of a company. You are not donating to an index. You are deepening one you also use.
For the material only you hold.
Plenty of what an organisation knows is not on a page anyone can reach. It sits in internal documentation, product manuals, corporate sites behind a staging domain, and the knowledge clusters a team has built up over years.
Indexer OS is an appliance, not an operating system you use. It turns hardware you already have into a machine that reads that material where it sits and keeps it current — and it asks nothing of you after the first boot.
Boot it once
It finds the network, works out what the hardware can do, names itself and joins the fleet. There is nothing to configure and nothing to patch.
Point it at what you hold
Corporate pages, public-facing sites, product documentation and the knowledge clusters your teams maintain. Indexer OS reads your own pages and files in place, rather than asking you to export them somewhere first.
It looks after itself
Machines lose power and get unplugged mid-sentence. The appliance expects that: work in progress is picked up again, and it updates itself for as long as it runs.
Spare hardware, with something to do
An old desktop, a server with room on it, a machine between projects. Capacity grows by adding a machine rather than by moving up somebody else's plan.
Most of what an organisation knows is a document.
A standard is a PDF. A manual is a Word file. A dataset is a spreadsheet and a board pack is a deck. None of those can be read by a browser the way a page can, so they have to reach a machine that can convert them — which is exactly what the appliance is for.
Word, PowerPoint, Excel and their OpenDocument equivalents, PDF, RTF, EPub, CSV and TSV, JSON and XML, and archived mail. Each is converted to readable text on the machine doing the reading and then read through, not skimmed for a summary. Images, video, audio and archives are skipped, because there is no text in them to read.
The machines already doing this.
A node you run joins these. The figures are read from the pool as this page loads.
One appliance, two kinds of shelf.
Indexer OS does not only feed the public index. The same machine can build collections that are yours alone — and a collection is a shelf, so it can be split the way your organisation is split.
Public collections
What you publish joins a shelf anybody can search. You can supplement a collection that already exists, or bring enough of something new that it becomes one.
- Supplement a shelf that is already there — add your specifications to the standards collection, your open-access papers to research
- Or seed a new one, where the material is a kind the index does not hold yet
- Answers quote your passage and carry your page back to the reader
What that looks like
- A standards body putting its normative text where the clause can be found, not just the document
- A university publishing open-access research and the data behind it
- A regulator's decisions, filed as judgments with their jurisdiction and dates
- A manufacturer's product manuals, so the fault code has an answer
Private collections
The same appliance, reading the material you would never publish. It becomes a shelf that is yours: not in public search, not in anybody else's answers.
- A shelf public search never reaches, and no other organisation can query
- Split it as your organisation is split — one per department, per team, per programme
- Where the material needs it, the storage is separated too, the way the health workspace already runs
What that looks like
- Engineering: runbooks, architecture decisions and incident write-ups, answerable without asking the person who wrote them
- Legal: negotiated positions and contract precedent, kept on its own shelf rather than in a shared drive
- Support: every resolved ticket and internal note, so the second person to hit a problem finds the first
- Research and development: lab notes and internal reports, on a shelf that never leaves the department
The examples are illustrative — shapes of organisation written to show what each kind of shelf is for. They are not customers, and nothing here is a reference.
Where this is going
Not built yetA partner running Indexer OS grows the index — their reading adds records to public collections and keeps them current. We are working towards giving that back as credit against askFinz usage, earned monthly from what a partner actually contributed.
None of it exists today, and we would rather say so than let it read as a feature. The hard part is the counting, not the credit: what counts as a contribution, how quality is judged, and what the same page read by two partners is worth. When there is something to quote, it will be on this page with a number attached.
A collection is not a filing cabinet.
Contributing is not archiving. Once material is on a shelf it is read from the workspaces people already work in — a question in Research, a company view in Finance, a course in Learn — so the answer arrives inside the task rather than being something somebody has to go and look up first.
Research
Research reads collections directly. A question is answered from the passage that addresses it, with the source attached — including your material, once it is on a shelf.
Finance
Finance reads the filings and sector shelves when somebody is building a view of a company, instead of asking them to go and find the document first.
Learn
Learn reads this material alongside course content brought in by the people using it, so a citation in a lesson points at the source rather than a paraphrase.
Chat, Search, Code, Projects, Storage
The apps page says which are live, which are in beta and which are planned — and keeps saying it, so it is worth reading there rather than being promised here.
Four steps, and one of them is ours.
- 01
Tell us what you hold
A couple of sentences: what you publish or keep, and what you would want read.
- 02
We read one back to you
We look at the material properly and come back with what we would do with it, and what we would not.
- 03
Reading starts
From the pages you publish for search engines, at a pace that stays out of your way — or from Indexer OS on your own hardware.
- 04
It becomes answerable
Pages become answerable as they are read. You can see your domain on the live status page.
Tell us what you hold.
There is no form-filling exercise behind this and no sales sequence. It sends one email to a mailbox a person reads, and you get a reply from that person.
- One email, to partners@askfinz.com
- A reply within two working days
- No mailing list, and your details go nowhere else
Only want a site read? You do not need us for that — put it in the queue and nobody has to be emailed at all.
The four we are always asked.
- What does this cost?
- Nothing to be read. If a partnership needs something built — a feed shaped a particular way, access to material that is not public — that is a conversation, and it starts with the form below.
- Do we have to give you our data?
- No. Public pages are read the way any visitor sees them. Material behind a login is read only where you have asked for it and supplied your own access, and Indexer OS keeps the reading on hardware you own.
- Can we stop?
- Yes, and quickly. Say so and reading stops. Nothing about this arrangement is designed to be difficult to leave.
- What if we are not a publisher?
- Most partners are not. Organisations with their own documentation, universities, standards bodies and platforms building on the index are all in scope — pick the closest option on the form and tell us the rest.