Built in retail.Not built for.

The Agentic Operating System for Retail Decisions. It watches the business overnight, explains what moved, proposes the fix, and writes the approved action back into your systems.

Watch a decision get made

OODIE · Morning briefWorked exampleMON 08:41

Three things need you this morning.

Cluster 3 is 6.8pp under planAct
Move CL-03 to markdown ladder B. Shallower first cut, held two weeks, size curve rebalanced from CL-04 transfers.
Expected GM$
+$61,400
Units at risk
2,140
Reversible
Yes, 14 days
A new tariff landed in the autumn buyOTB and margin impact are rebuilt.Review
Size curve broke in 41% of doorsThree pack changes prevent the next stockout.Resolve

Your systems

By default, nothing is written back until someone approves.

Price conditions, SAP S/4HANA, 214 SKUsWaiting
Markdown ladder B, CL-03, 18 doorsWaiting
Transfer orders, CL-04 → CL-03, 3 packsWaiting
OTB and MFP, re-projected for the seasonWaiting
Decision ledger, entry 4417Waiting

+4–8%
Gross-margin dollars
Same inventory, more margin.
−8–12pp
Discount depth
Shallower cuts. Units still cleared.
+1–3pp
Margin rate
Less reflex discounting in a busy week.
8–12 wks
To your first live cycle
Not a two-year programme.

Ranges measured in pilots and in production at ALDO Group, against an agreed control.

The people who use it

Five jobs. Five Mondays. Pick one.

Each role works with its own specialist agent, using the same numbers, the same evidence, and the same approval trail.

Choose a role to see their Monday.

MayaPlanner

Before OODARIS

“I used to rebuild the same forecast in three spreadsheets just to answer one question in the Monday meeting.”

With OODARIS

The variance is already explained when she opens her laptop. She spends the morning deciding, not reconciling.

Merchandise planning workspace with a laptop, calendar, and markers

MFPFinancial targets, 220+ KPIsMON 08:41

Cluster 3 is 6.8pp under planCause found. The fix is ready for approval.Act

Markdown depth ran deeper than the ladder plan61%
Traffic down across the cluster, week over week24%
Size-curve break on core generics15%

87% confidence, 14 sources

DanielBuyer

Before OODARIS

“We'd walk into the line review with last week's numbers. By then the risk had usually moved.”

With OODARIS

When a cost changes, OTB and the margin by department are rebuilt before the next purchase order goes out.

OTBBuy budgets, auto rebuildMON 08:41

A new tariff landed in the autumn buyOTB and margin impact are rebuilt.Review

Expected GM$Units at risk

01FW26-114+$38,200420

02FW26-087+$21,600610

03FW26-152+$9,3001,110

NiaAllocation

Before OODARIS

“By the time I saw a broken size curve, the best doors were out of core sizes and the slow doors were full.”

With OODARIS

Breaks are flagged door by door, each with a pack change or transfer ready to approve.

AllocationPacks, reflows, transfersMON 08:41

Size curve broke in 41% of doorsThree pack changes prevent the next stockout.Resolve

CL-04 transfer sequenced, 3 packs
Transfer orders raised, CL-04 → CL-03, 3 packs

TomFinance

Before OODARIS

“The variance showed up every month-end. Nobody could tell me which decision caused it, or what would close it.”

With OODARIS

Every approved change carries its expected P&L effect, so a variance traces back to the decision behind it.

Decision ledger, entry 4417, CL-03W+2

Simulated GM$: +$61,400
Actual GM$: +$64,910
Forecast error: −5.4%

Merchandise PlanningOwns the call, approved 08:47Approved
Merch FinanceInside the margin guardrail, auto-clearedCleared

Ledger entry 4417 sealed, reversible for 14 days

Composite roles drawn from live retail engagements

One real decision, end to end

This is what “agentic” looks like on a Monday.

01 / 07 · Monitor

It ran 1,412 checks before anyone logged in.

Sell-through, stock, traffic, price and plan, across every cluster, every night. Three things looked wrong enough to bother a human with.

Decision evidence1,412 checks, 3 exceptions

ClusterDoorsGM$ WTDGM%Δ Plan

CL-01 Flagship12$412,80674.1%+2.4pp

CL-02 Urban21$688,41473.8%+1.1pp

CL-03 Suburban18$505,11270.9%−6.8pp

03 / 07 · Recommend

It proposes. You decide. That order never changes.

Written the way a planner writes to a colleague: the action, the reason, what it should be worth.

Decision evidence+$61,400 expected GM$, reversible for 14 days

Recommended action

Move CL-03 to markdown ladder B. Shallower first cut, held two weeks, size curve rebalanced from CL-04 transfers.

Expected GM$+$61,400

Units at risk2,140

ReversibleYes, 14 days

Nothing downstream moves until someone clicks this.

04 / 07 · Coordinate, Execute

It acted: four functions in eleven minutes, then written back into your systems.

Everyone affected saw the same recommendation and evidence, their guardrails were checked automatically, and the approved changes were written back with an audit trail and a rollback that works.

Decision evidence4 functions, 11 minutes, 214 SKUs, 18 doors

MP

Merchandise PlanningOwns the call, approved 08:47

✓ Approved

MF

Merch FinanceInside the margin guardrail, auto-cleared

✓ Cleared

AL

AllocationCL-04 transfer sequenced, 3 packs

✓ Accepted

SO

Store Operations18 doors notified, signage task raised

→ Notified

Written back

Price conditions written, SAP S/4HANA, 214 SKUs

OK

Markdown ladder B applied, CL-03, 18 doors

OK

Transfer orders raised, CL-04 → CL-03, 3 packs

OK

05 / 07 · Simulate, Learn

Two weeks later, it checks its prediction against what happened.

Actuals come back against the simulation, and the ladder it didn't choose is costed too. The gap is the lesson.

Decision evidence+$64,910 actual, +$61,400 simulated, −5.4% forecast error

Two weeks later: actuals

Simulated GM$+$61,400

Actual GM$+$64,910

Forecast error−5.4%

What would have happened: simulated

Ladder A: What would have happened

GM$ end of season$1.42M

Average discount38%

Sell-through wk 1291%

Ladder B: Recommended

GM$ end of season$1.48M

Average discount30%

Sell-through wk 1290%

Same units cleared, 8 points less discount, $61K more margin

06 / 07 · Adjust

It adjusts what it got wrong, then re-checks every guardrail.

The plan was re-run with the updated weights against this morning's sales.

Decision evidenceWeights updated, 4 of 4 guardrails pass

What it learnedClearance elasticity in suburban doors was stronger than the model assumed.

Weights updatedCL-03 and two similar clusters

✓ Updated

Margin floorCL-03

✓ Pass

Discount ceilingladder B

✓ Pass

Stock coverCL-04 donor doors

✓ Pass

Size-curve floorcore generics

✓ Pass

07 / 07 · Publish

It serves the new version to every system that uses it.

The adjusted plan and model go out through the same APIs and integrations, so tomorrow's checks start from what it learned.

Decision evidenceNew version live, next Observe starts here

Updated weights published to the forecast serviceOK

OTB and MFP re-projected, season to dateOK

Reports and dashboards refreshedOK

Ledger entry 4417 sealed, reversible for 14 daysSIGNED

The next Observe starts from here.

You set the rules.

Every proposal is checked against your guardrails before anyone sees it. Move one and watch tonight's list change.

Tonight's proposals

Overnight run, 1,412 checks

Across: the deepest discount a proposal applies. Up: the gross-margin rate it leaves. Shaded: what your margin floor and discount limit allow.

3 need approval, 0 cleared on their own, 3 blocked

Guardrails

Margin floor68.0%

Deepest discount40%

Weeks of cover4 wks

Clears on its own underNever

Keep the size curve

Markdown ladder B for CL-03
Maya · Planning
+$61,400Needs approval
Rebuild the autumn OTB for the new tariff
Daniel · Buying
+$18,900Needs approval
Move 3 packs from CL-04 to CL-03
Nia · Allocation
+$4,800Needs approval
Pull core sizes from six slow doors
Nia · Allocation
+$7,300Blocked: weeks of cover
Match a competitor's price on 12 core styles
Daniel · Buying
+$9,600Blocked: margin floor
Clear 40 end-of-life styles at 55% off
Aisha · Assortment
+$3,100Blocked: deepest discount

How approvals work

Illustrative: six proposals from one night, simplified. Your guardrails come from your own plan.

From leakage to a governed loop.

MFPFinancial targets, 220+ KPIs

OTBBuy budgets, auto rebuild

AssortmentOptions, clusters, size curve

AllocationPacks, reflows, transfers

Pricing & MarkdownDepth, timing, tested

Manual translation

Reconciled by hand

Old rules dominate

Reacting too late

The P&L floor: margin gone permanently

The Agentic Loop

Actuals feed back

From leakage to a governed loop.

MFPFinancial targets, 220+ KPIsManual translation
OTBBuy budgets, auto rebuildReconciled by hand
AssortmentOptions, clusters, size curveOld rules dominate
AllocationPacks, reflows, transfersReacting too late
Pricing & MarkdownDepth, timing, tested

Planning handoffs are leaking margin.

Agentic AI for retail planning, in production

ALDO Group backs OODARIS and runs MFP and OTB on it. Five capabilities are available to new clients today, and five more are coming soon.

Read the ALDO Group story See all ten capabilities

Our merchandise financial planning used to live in spreadsheets and month-end. Now MFP and the AI run in the same loop. Targets re-project as actuals land, and the legacy planning system was retired in July 2026.

ALDO Group · Merchandise planning

Plan
- Demand forecasting
- MFP & OTB
- Assortment planning & optimization
Buy
- Purchase orders & procurement
- Sizing & prepacking
Flow
- Allocation planning & optimization
- Replenishment
- Transfers
Sell
- Pricing, markdown & promotions
- Order fulfillment

Bring one decision. We'll show you the loop around it.

See your decision mapped live.

A working session with your planning and merchandising leads. Bring one decision. We map who touches it and the data it needs, then show how OODARIS would check it, propose the fix and write the approved change back.

Book a demo

Run one decision live.

Pick the one decision where margin leaks first. Your team works it live in OODARIS, with named approvals and controlled writeback. Plan on 8–12 weeks to the first live cycle, then read the result against an agreed control and decide whether to scale.

Plan a pilot