AI-driven inventory optimisation for beauty, cosmetics and pharma distribution

Forecast demand, automate replenishment and keep every store balanced

IRIS forecasts sales for every SKU in every point of sale, computes the orders and sends them to your ERP. Your teams validate, IRIS does the rest.

Book a demo How it works WhatsApp
  • Live in 2 to 4 weeks
  • Connects to Sage, Odoo, Dynamics 365, SAP
  • EN · FR · AR support in your time zone
AI planning platform for beauty & pharma retail

Trusted by beauty and pharma distributors in Morocco and the region

  • Client logo
  • Client logo
  • Client logo
  • Client logo
  • Client logo
  • Client logo

The platform

The IRIS AI retail planning platform

Five modules that share one forecast, one stock picture and one connection to your ERP.

IRIS operation dashboard showing forecasts, order management, inventory and sales alerts

Inside the product

One platform, from the morning dashboard to the order in your ERP

Real screens from IRIS, connected to a 22-store cosmetics network. Every number a planner needs is one click from the order it should trigger.

01

Start the day with what needs a decision

The operations dashboard groups forecasts, orders, inventory and sales into three levels: immediate action, watch out, under control. Each line opens the SKUs behind it.

  • Anomalies ranked by business impact, not by alphabetical order

  • Filters by brand, category, city and store

  • Missing forecasts, stuck orders and active stock-outs in one view

Start the day with what needs a decision
02

Forecasts that know your calendar

Seasonal profiles are learned per category and per store, with Ramadan and Eid aligned on the lunar calendar. Fragrance, skincare and make-up do not peak on the same week, and IRIS does not pretend they do.

  • Profiles per category: fragrance, skincare, make-up, general retail

  • Year-on-year comparison and manual override with a reason

  • Launch ramp for new references without history

Forecasts that know your calendar
03

Validate, then let the ERP do the typing

Order proposals arrive per store and per supplier with the quantity, the value and the reason. Approve, adjust or reject; approved orders are created in Sage, Odoo, Dynamics 365 or SAP with the IRIS reference.

  • Approval thresholds by category and by value

  • Case packs, minimums and supplier calendars already applied

  • Status returned from the ERP: created, prepared, shipped, received

Validate, then let the ERP do the typing
04

See stock health before the write-off

Stock by lifecycle, by ABC class and by store, with critical and warning counts. Slow movers, expiring lots and overstock are visible weeks before they become a loss.

  • Red, amber and green per SKU and store

  • Stock value and coverage in days

  • Transfer proposals to the store that will sell in time

See stock health before the write-off

Client outcomes

Business outcomes within the first six months of IRIS

Target objectives observed on the pilot and defined in the product specification.

-40 to 50 %

fewer stock-outs on best-sellers

-30 to 50 %

less overstock and expiry loss

60 to 80 %

less planning time

< 6 months

return on investment

How it works

From your ERP to the shelf, in four steps

  1. Connect

    Your ERP feeds IRIS several times a day: items, stock by warehouse and transit warehouse, invoiced sales, supplier purchase orders. Sage 100cloud, Sage X3, Odoo, Microsoft Dynamics 365, SAP Business One, Oracle NetSuite, or scheduled CSV / SFTP imports.

  2. Forecast

    A machine-learning engine (Prophet, LightGBM, ARIMA, ETS, XGBoost) selects the best model per SKU and per store. Seasonal profiles, Ramadan / Eid / holiday calendar, cannibalisation and launches are built in.

  3. Propose

    IRIS computes the need per store: coverage, safety stock, supplier lead time, case pack, min / max. Stock already in transit is deducted.

  4. Validate & transmit

    The planner validates, with configurable approval thresholds. The order is created in your ERP with the IRIS reference, and the status comes back: created, prepared, shipped, received.

IRIS order validation screen listing proposals per store with approve and reject actions

Who it is for

Built for the people who run a distribution network

IRIS is designed for mid-market distributors of international beauty and pharma brands: 10 to 100 points of sale, 1,000 to 50,000 SKUs, 5 to 50 brands.

CEO and owner

Fewer stock-outs, less cash in the back office

One number per week: service level up, days of stock down, and orders that no longer depend on one person's memory.

  • Service level and stock value on one dashboard

  • ROI targeted within six months

  • No multi-year lock-in

Supply chain director

A forecast you can defend

Model accuracy per SKU and per store, supplier lead times measured from real deliveries, and every proposal traceable to its inputs.

  • MAPE and bias tracked per reference

  • Lead time and service level from receipts, not contracts

  • What-if scenarios before changing a rule

Purchasing and planners

Validate in minutes, not in Excel all morning

Proposals arrive computed, with case packs and minimums applied. The planner decides, IRIS writes the order into the ERP.

  • Approval thresholds you control

  • Testers and gifts with purchase on the same order

  • Full history of who validated what

Retail operations and store managers

Every store sees only its own stock and orders

Managers follow their coverage, expected deliveries and transfer proposals without touching the network view.

  • Scopes by store, brand and category

  • Alerts for stock-outs and expiring lots

  • Mobile access on the Enterprise plan

Industries

Built for cosmetics and pharma. Not adapted after the fact.

Perfume, skincare, make-up, haircare, parapharmacy and OTC. Each category has rules a generic retail tool does not know.

Beauty & cosmetics

Multi-brand, multi-store, shade by shade

Explore

Perfume & luxury

Selective distribution, luxury thresholds, launches and gift sets

Explore

Pharma & parapharmacy

Lots, expiry dates, OTC seasonality and automatic replenishment

Explore

See all industries

Case studies

Results from real store networks

All case studies

Before and after

What changes when ordering leaves the spreadsheet

Today, in Excel and the ERPWith IRIS
ForecastLast year's weeks copied forward, adjusted by feelOne model per SKU and store, chosen automatically, aligned on Ramadan and Eid
Order preparationStock export on Monday, orders built by WednesdayProposals refreshed with every ERP import, validated in minutes
Order entryRetyped by hand into Sage or OdooCreated in the ERP with the IRIS reference, status returned
Testers and gifts with purchaseConsumed without allocation, counted as stockAllocated at validation, tracked separately
ExpiryDiscovered at the write-offAlerts and transfer proposals weeks before the lot expires
Supplier lead timeThe contractual figure, never measuredComputed from real deliveries, per supplier
ReportingAssembled by hand for the Monday meetingService level, accuracy and stock health, live, per store

Why IRIS

Reasons distributors choose IRIS

  • Purpose-built for beauty and pharma

    Testers, gifts with purchase, shades, lots and expiry dates are first-class objects, not workarounds.

  • Two-way ERP integration

    Reads items, stock and sales; writes purchase orders back with the IRIS reference. No double entry.

  • Live in 2 to 4 weeks

    Connection, calibration on your history, parallel run, go-live. A free 30-day pilot on one or two stores.

  • Support in your time zone

    French, English and Arabic support, on-site training in Dubai and across the region.

  • Data hosted in Europe

    OVH data centres, one database per client, roles, MFA, SSO and audit log. Client-dedicated instance available.

  • No multi-year lock-in

    Monthly or annual contracts, export of all your data in CSV at any time.

50 000

SKUs handled per network

100+

stores per network

6

ERP connectors

3

languages, with RTL

Platform capacity as specified. Client figures are published only after validation.

Implementation

Live in four weeks, not four quarters

The same plan on every project. Your team spends a few hours a week; IRISYS does the rest.

  1. Week 1

    Connect and check

    ERP connector or CSV / SFTP export. Item master, stock by warehouse, sales history, open purchase orders. Data quality report.

  2. Week 2

    Calibrate

    Models trained on 12 to 24 months of history. Accuracy reviewed by category with the planner. Rules set: safety stock, service-level targets, case packs.

  3. Week 3

    Run in parallel

    IRIS proposals compared every morning with the current process. Differences explained; rules and events added.

  4. Week 4

    Go live

    Proposals validated in IRIS and created in the ERP. Thresholds opened category by category. On-site or remote training.

About IRISYS

A planning company for considered categories

IRISYS builds AI planning software for the supply chains behind beauty, luxury and pharmacy. IRIS is deployed with a luxury cosmetics distributor in Morocco and designed for the Maghreb, the Gulf and French-speaking Africa.

Our story

From the blog

Practical reading for supply chain teams

-20%
Forecasting

How to forecast Ramadan and Eid demand for cosmetics

Ramadan moves eleven days earlier every year. A forecast that copies last year's weeks will be wrong twice: before the month and after Eid. Here is how to align history, which categories move, and how to order.

6 min read

Read the article
Replenishment

From Excel to validated order proposals: a four-week plan

Moving a 20-store network from spreadsheet ordering to validated proposals does not take a year. It takes four weeks, a clean ERP export and one planner who wants it to work. Here is the plan we run.

7 min read

Read the article
Beauty

Testers and gifts with purchase: stop losing them in stock

Every launch consumes testers. Every campaign ships gifts with purchase. In most networks nobody knows how many, where, or whether they were sold by mistake. Here is what it costs and the rules that fix it.

5 min read

Read the article
Pharma

Lot and expiry management for parapharmacy chains

Lots expire in one pharmacy while the same reference is out of stock two pharmacies away. The fix is not more stock. It is stock by lot, an alert before expiry, and a transfer proposal in time.

6 min read

Read the article

All articles

FAQ

What supply chain directors ask us first

Ready to see your own numbers in IRIS?

Send us an anonymised sales export. We show you your forecasts and order proposals within 10 days.

Book a demo Chat with IRIS customer support