The benefits assistant, explained.
Your benefits advisor has set up a way for your staff to ask questions about their own coverage and get an answer straight from your plan booklet. This page is for whoever has to decide where to put it and answer questions about it — usually HR, sometimes the office manager, occasionally IT.
You do not need an account. The link needs nothing installed; the Slack option, if your advisor offered you one, needs a one-time approval by whoever administers your workspace.
What it actually is
A private web page, specific to your plan. An employee opens it, types a question in ordinary words — “how much physio do I get?” — and gets an answer taken from your own benefits booklet, with the page number it came from so they can check it.
It answers in whatever language the question is asked in. Someone can write in French, Punjabi, Tagalog or Mandarin and get an answer back in the same language, from the same booklet.
If your plan has more than one group — say management and everyone else, on different coverage — your advisor will have given you a separate link for each group. Those are not interchangeable. Sending the wrong one shows people another group's coverage, which is usually the one thing HR does not want circulating.
How to roll it out
- 1
Check the link yourself first
Open it and ask two or three things you already know the answer to. You will learn its voice in about a minute, and you will spot immediately if you have been sent the wrong group's link.
- 2
Put it where people already look
The intranet benefits page, the onboarding pack, a pinned message in whatever channel your staff actually read. It does not need its own announcement, though one helps.
- 3
Say what it is for
Something like: “Quick questions about what our plan covers — this answers from the booklet, any time, and nobody sees what you ask.” Setting that expectation is most of the battle.
- 4
Tell people where it ends
It explains the plan. It cannot tell anyone whether a specific claim will be paid, look up their claims history, or change anything. Those still go to the insurer or to you.
And what it will not
It answers
- What is covered, and up to how much
- Whether a referral or doctor's note is needed
- When maximums reset
- Who is eligible — spouses, children, students
- Coverage during leave
- How to submit a claim, and who the insurer is
It refuses
- Whether a particular claim will be paid — only the insurer decides
- Anything about an individual's claims or medical history
- Medical advice of any kind
- What the plan costs the company — premiums and renewals are your business, not your staff's, so it declines and points them to you
- Guessing. If the booklet does not say, it says so
It is an AI assistant, so it can be wrong. Every answer says so, and cites the booklet page it came from. Your official plan documents and your insurer are always the final word — that is worth repeating in whatever announcement you send.
Can you see what people ask?
No. Not you, not your benefits advisor, and not blankit. This is the first thing an employee wants to know before typing a question about, say, mental health coverage, and it is worth answering plainly when you announce it.
Questions are not kept at all. A message is sent to the AI service that works out the answer and is not stored afterwards. There is no transcript for anyone to read later.
What is kept is deliberately dull: a record that a conversation happened, roughly when, which booklet it was about, and a broad subject label like “dental” or “prescriptions” — so your advisor can see which topics keep coming up and fix the underlying confusion. Never a name, never the words anyone typed.
Your staff are told this themselves, in their own language, before they ask their first question. The full notice is at blankit.ca/legal/assistant, and your advisor's own privacy page carries the same text with their named privacy officer on it.
One thing to weigh first
The assistant can also be installed into your own Slack workspace, so staff can ask without leaving it. Your advisor will send an install link that someone who administers your Slack has to approve.
Before you do, know the trade. Slack is less private than the link. The workspace is yours, so what your administrators can see or export from it is governed by your Slack plan and your own policies — not by blankit, and not by your advisor. On some Slack plans that includes direct messages with apps.
We tell your employees that before their first answer, and point them back to the web link if they would rather ask where nobody at their employer could ever look. We would rather they hear it from us than assume otherwise.
It is also a slightly plainer assistant in Slack: English only, and without the page citations, because there is no booklet viewer there to jump into.
A reasonable default
Share the link first. If your staff keep asking in Slack anyway, add the Slack assistant then — you will know it is worth it, and you will have the privacy conversation with actual usage behind you.
Ask the person who sent you the link
Your benefits advisor set this up and can change it — which booklet a group gets, turning it off, or switching it on for a new plan year. They are the right first call.
For questions about the assistant itself — how it handles information, or how to have something corrected or deleted — your advisor's privacy page names the person accountable for it. Everything published about how it works is at blankit.ca/benefits-assistant and /legal/assistant.