QashierGM Is Live. The Catch: It Stays in Qashier
Qashier launched QashierGM on 22 October 2026 and says it is available at no additional cost to all Qashier merchants in Singapore through QashierHQ on web, Android and iOS. It is not a general manager for every tool in your company; today, it is an operating assistant for the data and actions Qashier already controls.
Look at the boundary, not the job title. An owner can ask why a payout is lower than sales, compare outlets, spot a weak product or prepare a promotion without assembling reports by hand. But a supplier email, a meeting brief or work living outside Qashier remains outside that operating picture unless a promised connector eventually covers it.
Key takeaways
- It is live in Singapore now: QashierGM is available to all of its Singapore merchants at no additional cost as of October 2026.
- Its best answers come from operational context: It reads the merchant’s own sales, payouts, menu and outlet records instead of waiting for someone to paste a report into a general-purpose chatbot.
- Actions still stop for approval: The product shows the proposed change before a price, promotion or sold-out status goes live.
- The boundary is the buying test: Businesses already centred on Qashier get the shortest path to value; businesses with work spread across other systems should map the uncovered job before switching.
What changed with QashierGM on 22 October?
The launch moves Qashier from dashboards and reports towards a conversational layer that can answer, analyse, advise and act on merchant operations.
| Job | What QashierGM does | Where a person stays involved |
|---|---|---|
| Understand a payout | Explains sales, fees and refunds line by line from Qashier data | Owner checks the underlying transactions and resolves exceptions |
| Compare outlets | Surfaces performance differences and the records behind them | Manager decides why the gap exists and what response is appropriate |
| Change a menu item | Prepares a sold-out or price update across selected channels | Authorised person reviews the scope and confirms before it goes live |
| Send an invoice in Crew | Creates the invoice and payment link from the team conversation | Role permissions apply; larger changes route to a manager |
| Watch for exceptions | Sends a digest or an alert such as unusual refunds | Owner investigates the pattern before acting on staff or customers |
The launch followed an open beta that began in September. The most common merchant questions concerned sales trends, best sellers and payouts. Those are good first jobs for an assistant because the answer should come from a traceable ledger of transactions, not from plausible-sounding text.
Why does staying inside Qashier help?
An AI assistant becomes useful when it can see the current record and prepare the next action. QashierGM starts with both: the sales and payout data already captured by the terminal, plus permission to prepare changes inside the same operating system. That removes the copying step that makes a blank chatbot feel clever in a demo but slow on Monday morning.
The approval design is equally important. Qashier’s product page says nothing changes until the user confirms, while Crew’s product page says it applies the permissions of the person who tagged the agent and routes bigger changes to a manager. The pattern to copy: “AI gathers the evidence and stages the change; an accountable person releases it.”
What is the QashierGM catch for Singapore SMEs?
The catch is scope. QashierGM is strongest where the business record already lives in Qashier: sales, payouts, products, outlets, invoices and selected team conversations. That is a compelling fit for an existing merchant whose daily questions sit inside those records. It is not yet evidence that the assistant can run the unrelated work scattered through every inbox, calendar, document folder and specialist system.
Before relying on work outside Qashier, ask which messages the agent can read, which actions it can prepare, who approves, what gets logged and how a mistaken action is reversed.
Should you use QashierGM or a broader AI assistant?
Choose the assistant around the job and its source of truth, not around the most ambitious job title.
| Situation | Best starting point | Reason |
|---|---|---|
| You already run sales, payouts and menu operations in Qashier | Test QashierGM first | It starts with live operational context and built-in approval paths |
| Your main pain is reading reports after closing | Test three recurring questions in QashierGM | A short trial can show whether answers are accurate and save manual report assembly |
| The job crosses email, calendar, documents and several systems | Map a broader assistant workflow | The assistant needs context and actions beyond one merchant platform |
| The process is low-volume or changes every time | Keep a checklist or handle it manually | Setup and monitoring may cost more attention than the job |
| The work needs judgement across many unrelated cases | Hire or retain an operations person | Human context switching and accountability are the core requirement |
A broader personal AI agent fits a different shape of work: preparing a meeting from messages and files, following up across conversations, or bringing the next decision to an executive in chat. A purpose-built merchant assistant is usually the better choice when the job begins and ends in the merchant platform. Ask which assistant can complete your recurring job with the fewest blind handoffs. Feature counts matter less.
How should a merchant test QashierGM?
- 1. Write down three questions you already ask weekly. Use real jobs such as explaining a payout, comparing outlets and finding unusual refunds. Record how long the manual answer takes and which source proves it.
- 2. Check the answer against the record. Open the underlying transactions. Test refunds, voids, split payments and days when settlement timing makes sales differ from payouts.
- 3. Test one reversible action. Prepare a sold-out update or a limited promotion, inspect exactly which outlets and channels will change, then approve only when the scope is correct.
- 4. Test staff and manager roles too. Ask a staff member and a manager to request the same change in Crew. Confirm that permissions and escalation match the way the business actually delegates.
- 5. Review data responsibility. Document purpose, access, retention, incident handling and offboarding before adding more customer or staff data.
- 6. Decide from completed work. Measure accurate questions answered, approved changes completed and exceptions caught. Do not count prompts sent or summaries produced as business outcomes.
What should Singapore SME owners do now?
If you already use Qashier, pick one week of real operating questions and run the test above. The product’s tight boundary may be exactly why it works: the data is already there, the action is close and approval is visible.
If you do not use Qashier, do not switch an operating system just to acquire an AI label. Map the job first. Compare staying manual, improving a spreadsheet, hiring an administrator, using the assistant inside your current system or building a workflow across the tools you already trust. The winning option is the one that closes the loop without hiding the approval.
Sources
Common questions
- What is QashierGM?
- QashierGM is an AI assistant inside QashierHQ that answers questions from a merchant’s sales, payout, menu and outlet data. It can also prepare selected actions, such as changing a menu item or creating a promotion, but the change waits for user confirmation.
- Is QashierGM available in Singapore?
- Yes. Qashier launched QashierGM in Singapore on 22 October 2026. It is available through QashierHQ on web, Android and iOS to all Qashier merchants in Singapore at no additional cost as of October 2026.
- Can QashierGM explain why a payout differs from sales?
- Yes. QashierGM can break down a payout using the merchant’s own sales, fees, refunds and settlement records. The merchant should still check the cited transactions and investigate exceptions. An AI explanation is not the accounting record.
- Does QashierGM make changes without approval?
- QashierGM shows what will change and waits for confirmation before actions such as a menu or price update. In Crew, the agent acts with the requesting user’s permissions and routes larger changes to a manager for approval.
- Who should choose QashierGM instead of a broader AI assistant?
- QashierGM is the natural first test for an existing Qashier merchant whose recurring job begins and ends with sales, payouts, products or outlets. A broader assistant is more suitable when the job must cross email, calendars, documents and several business systems.
