Flu Season Sick Calls: AI Shift Fill for Burnaby Teams

It's 5:40 a.m. on a Tuesday in November. Your line cook texts that they've got a fever. Ten minutes later your MOA at the clinic says the same thing. Now somebody has to scroll through a contact list, text people one by one, wait, and hope. That somebody is usually the owner.

This is a narrow, boring job, and that's exactly why it suits AI. Not to replace your manager, but to do the first 30 minutes of phone tag so a human only has to approve the result.

How big is the sick call problem, really?

Statistics Canada tracks this. In its Work absence of full-time employees by industry table (table 14-10-0191-01, released January 2026), full-time employees in health care and social assistance lost 16.8 days per worker in 2025, with 14.3 of those days due to illness or disability. In accommodation and food services the figure was 7.5 days per worker, 6.1 of them for illness or disability.

Two honest caveats. These are national numbers, not Burnaby numbers. And they only cover full-time employees, so they likely undercount restaurants, where a lot of staff are part-time. Still, the clinic number stands out: more than three working weeks a year per full-time worker.

In BC the rules also shape how you respond. According to the province's paid sick leave page, employees covered by the Employment Standards Act get "5 days of paid leave per year for any personal illness or injury" once they've worked for you for 90 days, on top of 3 unpaid sick days. That includes part-time and casual staff. So a sick call isn't a discipline issue to argue about. It's a gap to fill.

What an AI shift fill agent actually does

Think of it as a dispatcher that never sleeps. A typical setup looks like this:

1. The sick employee texts a dedicated number or uses a short form. The agent logs the call, thanks them, and doesn't ask why they're sick.
2. It checks the schedule for who is qualified for that role, isn't already working, and wouldn't be pushed into overtime you didn't approve.
3. It texts those people in an order you set, for example seniority, or whoever has picked up the fewest extra shifts this month.
4. The first person to reply YES gets a hold, not a confirmation. The manager gets a one line summary and taps approve.
5. Once approved, the agent confirms the replacement, tells everyone else the shift is filled, and updates the schedule.

Step 5 matters more than it looks. BC's minimum daily pay rule says "an employee who reports for work must be paid for at least 2 hours, even if they work less than 2 hours." If two people both think they got the shift and both show up, you're paying for the extra one. A good agent closes the loop fast so that doesn't happen.

Why this suits AI (and where plain automation is enough)

Honestly, a lot of this is just rules: who's qualified, who's available, who's next. You don't need a language model to sort a list.

Where AI earns its place is the messy human part. People reply "can do after 11" or "only if I can leave by 3" or "who is this?" A rules engine chokes on those. A language model can read them, check them against the open shift, and either make a sensible counter offer or flag it for the manager. It can also answer the follow up questions staff always ask, like where to park or which entrance to use, from your own notes.

If your replacements always reply with a clean YES or NO, a simple automated text blast from your scheduling app may be all you need. That's cheaper and has fewer moving parts.

Restaurants and clinics need different guardrails

For a restaurant, the main checks are role and station. A dishwasher can't cover the grill unless you say they can. Keep the qualifications list current or the agent will happily text the wrong people.

For a clinic, the bar is higher. Some roles are regulated, and the agent has no way of verifying a licence on its own. It should only offer clinical shifts to people a human has already marked as cleared for that role. Front desk and MOA coverage is the sweet spot. Covering a nurse or a physician is a job for your existing processes.

Privacy: what BC law expects

Shift data is employee personal information. The Office of the Information and Privacy Commissioner's Guide to B.C.'s Personal Information Protection Act (2015) says your collection and use of employee information "must be reasonable for the purpose of starting, managing or ending an employment relationship," that you must notify employees when you collect it without consent, and that if you make a decision about an employee using their information, you must keep that information for one year.

In practice that means:

• Tell staff in writing that an AI agent handles sick calls and shift offers, and what it stores.
• Don't collect symptoms or diagnoses. "Sick, can't come in" is enough to fill a shift. BC lets employers request "reasonably sufficient proof of illness," but that's a manager conversation, not a chatbot one.
• Keep the shift offer and acceptance log for at least a year, since who got offered extra hours is the kind of decision people later question.
• Only give the agent the data it needs: names, roles, availability, phone numbers. Not payroll, not performance notes.

Canada's federal, provincial and territorial privacy commissioners said much the same thing about AI in their joint Principles for responsible, trustworthy and privacy-protective generative AI technologies (December 2023). One line worth pinning above your desk: "Accountability for decisions rests with the organization, and not with any kind of automated system used to support the decision-making process." That's why the design above keeps a human approving each fill.

Where this doesn't apply

This isn't for everyone, and it's worth being blunt about it.

• Small, tight teams. If you have eight staff and one group chat where someone always says "I got it," an agent adds cost and very little speed.
• Unionized workplaces. If a collective agreement sets a call in order, the agent has to follow it exactly. Get that order confirmed in writing first, or don't automate it.
• Messy schedules. If your schedule lives in a manager's head or a photo of a whiteboard, the agent has nothing to read. Fix the schedule first.
• Clinical coverage. As above, regulated roles need human checks the agent can't do.
• No bench. If nobody is available, faster texting won't create people. The real fix there is hiring or a casual pool.

There's also no published, independent study we could find that measures how much time AI shift filling saves small restaurants or clinics specifically. Treat any vendor's savings claim, including ours, as something to test on your own schedule.

A quick, illustrative way to size it

Here's an illustrative example, not a client result. Say a Burnaby restaurant gets three sick calls a week in flu season, and each one costs the manager about 30 minutes of texting and calling. That's roughly 90 minutes a week, mostly at awkward hours. If an agent cuts that to a two minute approval, you get most of that time back and the shift gets filled sooner. Run your own numbers. If you get one sick call a month, the honest answer may be that you don't need this.

Getting started

Start small. Pick one location and one role, set the call order, write the staff notice, and run it alongside your current process for two weeks. Check every fill. Then decide.

If you run a restaurant or clinic in Burnaby, New Westminster or anywhere in Metro Vancouver and want to see whether a shift fill agent makes sense for your team before flu season peaks, book a free call with Autana. We'll look at your schedule and tell you straight if it's worth doing.

Sources

• Statistics Canada. "Work absence of full-time employees by industry, annual" (Table 14-10-0191-01). Released January 2026. https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1410019101
• Government of British Columbia. "Paid sick leave." Employment Standards, 2022. https://www2.gov.bc.ca/gov/content/employment-business/employment-standards-advice/paid-sick-leave
• Government of British Columbia. "Minimum daily pay." Employment Standards. https://www2.gov.bc.ca/gov/content/employment-business/employment-standards-advice/employment-standards/wages/minimum-daily-pay
• Office of the Information and Privacy Commissioner for British Columbia. "A Guide to B.C.'s Personal Information Protection Act for Businesses and Organizations." 2015. https://www.oipc.bc.ca/guidance-documents/1438
• Office of the Privacy Commissioner of Canada, with provincial and territorial privacy commissioners. "Principles for responsible, trustworthy and privacy-protective generative AI technologies." 2023. https://www.priv.gc.ca/en/privacy-topics/technology/artificial-intelligence/gd_principles_ai/