Drink builders that standardise every recipe, compliance enforced at the till, and the margin per pour visible while the shift is still running — 17 beverage solutions, narrated by Iris.
Bars and juice counters leak by the ounce: free pours, syrup drift, age-gate risk, weather nobody planned for.
Pour cost — what the drink actually costs once the free pour is counted.
Draught yield — what came out of the keg against what was sold from it.
Tab recovery — the tabs that closed, and the ones that walked.
Happy-hour return — whether the window paid for the discount it gave away.
Zentallio meters the flow — and the margin per pour is visible while the shift is still running.
Drink builders standardise every recipe, so a cocktail costs the same whoever is behind the bar.
Compliance is enforced at the till — the age gate and the service log are part of the transaction, not a policy on a wall.
Weather-driven demand shapes prep, because a warm Friday and a wet Tuesday are not the same business.
Pour cost, draught yield, tab recovery and happy-hour return — live, with Iris naming the shift where the spirits drifted.
Orders, builds, tabs and any tender on one screen — with the age gate and the service log inside the transaction rather than beside it.
The build is the recipe. A cocktail rung at the till depletes the exact spirits and syrups it specifies, so pour cost is a fact rather than an estimate.
Compliance can't be skipped. The age gate fires before the line is added, and the check is logged against the transaction.
Iris upsells at the bar. The premium spirit, the second round, the shisha add-on — timed to what's already on the tab.
Pour cost, draught yield, tab recovery and happy-hour return on one live board — with Iris naming the shift where the spirits drifted.
Every module feeds one board. Till, cellar, draught lines and workforce all post into the same financial view.
Visible during the shift. Margin per pour is readable while the shift is running, not reconstructed the morning after.
Iris names the shift. Not that pour cost rose — which shift, which line, which bartender.
P&L, balance sheet, cash and close on a single balanced ledger — sub-ledgers that reconcile, and a month-end that is a review rather than a rebuild.
Cellar value is a live number. Bottles, kegs and open stock carry their cost into the balance sheet without a manual count.
Unrecovered tabs are visible. A walked tab posts as what it is, rather than disappearing into shrinkage.
Ask Iris for any number. Any venue, any period — answered in the conversation, not a week after close.
Demand-planned replenishment across every venue. Iris forecasts by daypart and weather so the right stock lands before you need it — not after.
The weather drives the order. A forecast warm weekend becomes keg and mixer quantities, not a manager's hunch on Thursday.
Cellar and bar are one chain. Delivery, cellar, line and glass are a single movement rather than four separate records.
Feeds the pour ledger. Receipts post to Bottle, Wine & Cellar Inventory directly, so variance is caught same-day.
Digital checklists, shift routines and standard operating procedures across every venue — completion tracked live, exceptions escalated.
Line cleaning, signed off. The routines that decide draught quality are completed in the system, not assumed to have happened.
Cash-up and tab close, tracked. End-of-shift routines roll up by venue, so an outlier shift is visible the next morning.
Exceptions escalate themselves. The routine that didn't complete surfaces to a human; the rest stays quiet.
Rostering, attendance and labour percentage measured against forecast demand. Iris flags the overstaffed lulls and the understaffed peaks before the shift, not after.
Rostered to the weather. Bar cover is built from forecast demand, which in this sector means the forecast itself.
Happy hour is staffed as a peak. A pricing window is a demand event, and the roster treats it as one.
Certification tracked with the shift. Responsible-service qualifications are held against the person rostered, not in a folder.
Deterministic logic under the hood (the “Zen Rules” layer) — age gates, tabs, cellar counts and compliance logs behave identically every time, because in this sector a judgement call is a liability. Five solutions run on this layer.
Four tabs still open on the floor, one above threshold. Flagged to the duty manager while the guests are still in the room — not discovered at cash-up.
Age-restricted lines cannot be sold without a check. The gate fires at the till, the result is logged against the transaction, and the rule does not bend for a busy bar.
Any age-restricted product triggers the gate before the line can be added — it is part of the transaction, not a policy on the wall.
The check, the method and the staff member are recorded against the sale, so the evidence exists if it is ever asked for.
Thresholds are configured per market, because the legal age and the challenge policy are not the same everywhere.
A refused sale is logged as a refusal rather than a void, which is the record a regulator actually wants to see.
The same gate applies on app, kiosk and delivery channels, so the strictest surface is not the only compliant one.
A tab opened at the bar stays open until it is closed — with pre-authorisation, running totals, transfers and threshold alerts, so tab recovery stops being a hope.
A tab is opened against a card pre-authorisation, so the venue is not carrying the risk unsecured.
Rounds are added to the running tab without re-opening an order or duplicating a line.
A tab crossing a configurable threshold is flagged to the manager while the guests are still in the room.
Tabs transfer between the bar and a table, and split any way the group wants at settlement.
Tabs still open at last orders are surfaced as a closing task, so recovery happens before the floor empties.
Bottles, kegs and open stock tracked as real volume rather than as units on a shelf — depleting through the recipe as the till sells, with vintage and bin held where it matters.
An open bottle is tracked by remaining volume, so the system knows a spirit is running out without a physical count.
Wine is held by vintage and bin, so staff know availability before they recommend rather than after.
Every sale depletes the cellar through the drink's own recipe, making stock a consequence of trading rather than a weekly chore.
Par levels come from forecast demand and supplier lead time rather than habit, and move with the season.
Cellar value is carried live into the ledger, so what is on the shelf and what is on the balance sheet agree.
The record a licensing authority actually asks for — refusals, incidents, certifications and service decisions, logged as they happen rather than reconstructed afterwards.
Refusals of service are logged with time, reason and staff member, building the evidence trail as a by-product of the shift.
Incidents are captured on the floor rather than written up from memory at the end of the night.
Staff responsible-service certifications are held against the person and checked against the roster, with expiry flagged ahead of time.
Licensing hours are enforced at the till, so a sale outside permitted hours is prevented rather than explained.
The compliance pack is generated on demand, ready for inspection rather than assembled the night before one.
Every system a venue depends on, connected and watched: payments, the delivery aggregators, local tax and excise rules, and the accounting ledger — configured per market, not coded per customer.
Processors and payment methods are both covered — the guest pays with whatever they carry, you get one settlement file.
One menu publishes to every aggregator against one live stock, and their orders land in the same order book as the bar.
Tax and alcohol duty are configured per market, which is why a new country takes days rather than a release.
Sales, tax, tips, refunds and settlement post themselves to the accounting system, already coded to the right accounts.
Every connection is health-checked continuously — when an API degrades, Iris quantifies the orders at risk rather than letting a venue find out at the rush.
Forecasting, cost and anomaly models under the hood (the “Zen Models” layer) — weather-driven demand, keg yield, pour cost and happy-hour return, predicted before the shift rather than reviewed after it. Ten solutions run on this layer.
The order reaches the bartender as a build, not a shout. Tickets route by station and the queue is forecast rather than absorbed — because at a bar the queue is the revenue ceiling.
An order becomes a printed or screened build the moment it is rung, with every spec on it — no verbal relay across a loud room.
Tickets route by station, so draught, cocktails and bottles are not competing for one pair of hands.
A queue-wait estimator predicts the current wait, so staffing and service-well setup are decided against a number.
Rounds are sequenced to land together, because a table served in three waves is a complaint waiting to happen.
Drinks per hour by bartender are tracked, identifying where the bar actually bottlenecks rather than where it is assumed to.
Every drink carries a costed recipe, so margin per pour is a live number rather than a monthly discovery — and a free pour has somewhere to show up.
Each drink has one specification — measures, garnish and glass — and the platform costs from the same card the bartender builds from.
Cost per pour recomputes the moment a spirit or syrup price moves, and anything crossing its margin floor is flagged with the size of the gap.
Theoretical usage against actual is the number that catches the free pour, because generosity is invisible one drink at a time.
Batched and pre-mixed recipes nest inside drinks, so changing a syrup once recosts every cocktail built on it.
Menu engineering runs on real margin per pour rather than on which drinks the team enjoys making.
What came out of the keg against what was sold from it. Yield is measured per line, and the gap — foam, line loss, unrecorded pours — is raised as an anomaly rather than absorbed.
Expected yield per keg is compared to what the till actually sold, line by line, so the shortfall has a location.
Foam, line loss and cleaning waste are separated from unrecorded pours, because the fix for each is different.
A line drifting outside its normal yield band raises an anomaly while the keg is still on, not at the next stocktake.
Line cleaning schedules are tracked against yield, so quality maintenance is visible in the number it protects.
Keg changeover is logged automatically, giving a true picture of consumption rather than a count of empties.
Theoretical against actual across every liquid on the premises — separating spillage, breakage, comps, staff drinks and the slow over-pour that quietly moves pour cost.
What the recipes say should have been poured against what actually left the bottle. The gap is the number that decides pour cost.
Spillage, breakage, comps, staff drinks and trial pours are separated, because each has a completely different fix.
Slow over-pouring is caught by variance pattern long before it shows up as a margin problem in the P&L.
Variance is attributed to shift and bartender, so coaching is specific rather than a general reminder at the briefing.
Syrup and mixer drift is tracked alongside spirits, because the cheap ingredients are where the volume hides.
Few things move drink demand like the weather. Temperature, rainfall and daylight are model inputs rather than context a manager carries in their head — driving prep, stock and the roster.
Temperature, rainfall, humidity and daylight are first-class features of the demand forecast, not an afterthought.
Elasticity is learned per venue — a terrace and a basement bar respond to the same forecast very differently.
Category splits shift with the weather, so a warm evening changes the mix as well as the volume.
A material forecast revision during the day re-flags the affected prep and stock plans while they can still change.
One forecast feeds prep, ordering and rostering, so the three cannot contradict each other.
A pricing window is an investment, and it should be measured like one. Iris sets the window, the depth and the products — then reports whether the discount actually bought incremental trade.
Windows are configured by day, hour, venue and product group, and switch automatically across every channel at once.
Discount depth is modelled against expected uplift, so the window goes only as deep as it needs to.
Incremental trade is separated from trade that would have happened anyway — the only honest measure of a happy hour.
Margin is protected per line, so a window cannot quietly sell the highest-cost pour at the deepest discount.
The window ends on schedule everywhere, so nobody is manually reverting prices at the busiest moment of the evening.
Summer spritz gives way to winter warmer on schedule, across every channel in the same second — and which drinks earn their place is decided on forecast sell-through, not on affection.
Seasonal menus are configured once and switch simultaneously on till, app, kiosk and every aggregator.
Each menu slot is scored on forecast sell-through and margin, so the question becomes what a slot earns.
A new drink that would simply take sales from an existing one is flagged before it takes a place on the list.
Ingredient lead time is respected, so a menu never goes live before the stock to serve it has landed.
Limited-time runs carry a start, an expected volume and an end, rather than trailing off unnoticed.
A high-margin, high-dwell add-on with its own consumables, coal cycle and compliance profile — managed as its own service line rather than bolted onto the drinks menu.
Flavours, heads and coal are tracked as their own consumables, so the true margin on a session is visible.
Sessions are timed, with coal changes and refills prompted rather than left to the guest to chase.
Table dwell is forecast against the session, because a shisha table turns very differently from a drinks table.
Age and area restrictions attach to the line, so the same compliance discipline applies as to alcohol.
Demand is forecast alongside drinks, so coal and flavour stock are ordered against the same weekend curve.
Bar, table-side, app, web and the aggregators on one order book — same menu, same live stock, same ledger, and a promise time that accounts for the queue actually at the bar.
Every source writes into the same queue, so the bar works one sequenced list rather than a screen per channel.
A keg blowing at the bar removes the line from app, web and every aggregator in the same second.
Promise times are channel-aware, because table-side and delivery deserve different numbers rather than one optimistic one.
Table-side ordering carries the tab, so a round ordered from a phone joins the same open tab as one ordered at the bar.
Each channel settles into the same ledger with its commission and delivery cost attached, giving true margin by channel.
A regular who stops coming rarely announces it. The classifier identifies guests drifting out of their own pattern and proposes the offer most likely to bring each one back, while the habit is still recoverable.
Each guest is scored against their own visit rhythm, not a single estimate applied to everybody.
A guest sliding from weekly to monthly is flagged automatically — the segment change is the alert.
Seasonal pauses are distinguished from real churn, because a terrace regular in January has not left.
The optimiser picks the offer and channel with the highest modelled return for that guest, rather than a blanket discount.
Win-back outcomes are tracked, so the model learns which interventions actually work rather than which felt generous.
Not a screenshot — the actual agent. Natural-language answers over the live bar. Two solutions run on this layer.
Every way a drink can be ordered — base, size, ice, mixer, shots, syrups, milk, garnish — captured accurately, priced for what it costs, and standardised so the recipe holds whoever is pouring.
The build is the recipe: what the guest chooses is what the bartender makes and what the ledger costs.
Clustering surfaces the combinations guests actually order as one-tap presets, which matters most when the bar is three deep.
Every modifier carries its own price and its own cost, so a premium spirit or an extra shot is never absorbed silently.
Allergen and dietary flags travel with the build — milk alternatives, nut syrups, gluten — and incompatible combinations are blocked.
Non-alcoholic and juice builds run through the same engine, because a juice counter has exactly the same syrup drift problem.
A recurring pass turns an occasional visit into a habit with predictable revenue — sitting on the same wallet as points and streaks, on one balance across every channel.
A daily coffee or weekly pint pass is sold as a recurring product, with entitlement checked at the till in a single tap.
Pass, points and streaks share one wallet, so a guest never has to work out which scheme they are using.
Iris picks the reward with the highest modelled redemption for that guest, rather than a blanket discount on people already coming.
Pass economics are modelled honestly — redemption rate against price, so a popular pass is not quietly a loss-maker.
Deferred revenue and outstanding points are valued live and posted to finance rather than discovered at year end.
Seventeen solutions. Every card opens a live guided demo.
You configure a sector playbook, not a custom project. Iris applies it herself — agentically, from day one.
Onboarding is agentic. Iris connects the till, the cellar and every aggregator feed directly — no manual data mapping.
The playbook applies itself. Sector thresholds go live immediately, then tune from real trading data, venue by venue.
A new market takes days. Tax, duty and licensing rules are configured per market rather than coded per customer.
The first line of support is agentic — Iris resolves most of it herself. Our engineers pick up from there.
Layer 1 — Iris, 24/7. Configuration questions, anomalies and routine issues resolved directly, instantly.
Layer 2 — our engineers. Anything Iris can't close escalates automatically to a Zentallio engineer.
No blank tickets. Every escalation arrives with Iris's own diagnosis — engineers start from an answer.
“Bars and juice counters leak by the ounce: free pours, syrup drift, age-gate risk, weather nobody planned for. The product is liquid; the losses are too.”
Zentallio meters the flow: drink builders standardise every recipe, compliance is enforced at the till, weather-driven demand shapes prep, and the margin per pour is visible while the shift is still running. Underneath sits the same platform every sector runs — one data spine from the till to the ledger, three layers of intelligence, and an agent that narrates every screen, flags what needs a human, and acts on the rest.
Free pours and syrup drift — invisible one drink at a time, decisive across a shift.
Age-gate risk carried by a busy bartender's judgement instead of by the till.
Weather nobody planned for, and a happy hour that gave away margin it never earned back.
The age gate, the tab and the compliance log fire deterministically — the rule does not bend for a busy room.
Pour cost, keg yield, weather demand and happy-hour return are modelled before the shift, not after it.
“Which shift did the spirits drift on?” — answered from the same ledger the drink is costed from.
Every Zentallio sector runs the same three-layer intelligence, tuned to how that format actually loses margin.