Platform / Data flow

From scattered transactions
to one live picture.

Six steps. No replacement of your current systems. This is how raw data from every store becomes a single management surface owners and operators can actually run a network on.

Quick answer

marql reads POS, ERP, accounting, e-commerce, and CSV or API exports, maps them into one chain-wide data model, computes daily KPIs and anomalies, and turns that into role-based cockpits and tracked store actions.

Access model

Read-only by default

Sources

POS + ERP + e-commerce + CSV/API

Replacement

No rip-and-replace

What the flow produces

What owners see after the first sync.

This is not an integration diagram for its own sake. The flow produces a morning briefing, role-based views, and an action layer the network can actually operate from.

What we read

Transactions, stock movements, invoices, prices, schedules

What we standardize

SKU, store, margin, labor, supplier, and compliance data

What the team gets

Morning briefing, alerts, role-based views, checklists, tracked actions

Morning briefing

Today · 08:30

Day 1 briefing
598K €
Chain revenue · MTD
↑ 2.8% vs plan
↑ 2.5% vs 7d

Revenue vs plan target for the current month.

26,4%
Avg. margin
158K € total
↑ 0.4pp vs 7d

Gross margin across active stores. Review P&L if it keeps moving.

13,80 €
Avg. receipt
12 active stores
↑ 0.8% vs 7d

Average transaction value from connected POS data.

3,4
Items / order
avg SKU lines / receipt

Average SKU lines per receipt. Below 2 can signal missed upsell.

Step
01

Sources connect

ERP · POS · E-commerce · WMS · 1C · Excel

marql plugs into the systems you already run. No replacement, no rip-and-replace. Read-only connectors pull transactions, stock movements, price lists, schedules, suppliers and HR data from every store and channel.

Examples
  • SAP / Oracle / 1C
  • Square / iiko / R-Keeper
  • Shopify / WooCommerce
  • Custom ERP via API
Step
02

Data is normalized

One schema across the entire network

Every store calls a SKU something different. Every POS reports margin its own way. marql maps them into a single canonical model — products, stores, transactions, costs, headcount, suppliers — so cross-store comparisons actually mean something.

Examples
  • SKU mapping & deduplication
  • Currency & VAT handling
  • Cost & margin reconciliation
  • Time-zone alignment
Step
03

Layers are built

From raw events to management metrics

On top of the unified data, marql computes the layers a network operator actually uses: real-time KPIs, daily P&L per store, margin per category, stock health, labor cost ratios, franchise compliance scores.

Examples
  • Network KPI layer
  • Margin & loss layer
  • Operational tasks layer
  • Franchise governance layer
Step
04

Anomalies surface

Problems found before they hurt margin

Rules and statistical baselines flag what doesn't fit: a store dropping 18% week-over-week, shrinkage spike on a SKU, a manager skipping cash reconciliation, a franchisee out of compliance. Each anomaly comes with context, not just a red dot.

Examples
  • Margin drift alerts
  • Stock & shrinkage alerts
  • SLA / checklist breaches
  • Franchise audit flags
Step
05

Manager sees the result

One AI operator · every role · every store

The owner sees the network. The COO sees operations. The store manager sees today's tasks and their numbers. The franchise lead sees compliance. Same data, different views — no spreadsheets, no Monday-morning dashboard rebuilds.

Examples
  • Owner / CEO view
  • COO / Ops view
  • Store manager view
  • Franchise partner view
Step
06

Decisions go back to stores

Playbooks · standards · actions

Insight without action is a dashboard. marql pushes standardized playbooks, daily checklists, price updates and operational corrections back into the network — and tracks whether they actually got done.

Examples
  • Daily ops checklists
  • Standardized playbooks
  • Price & promo rollouts
  • Compliance follow-ups

Data Flow FAQ

Questions buyers ask before they grant data access.

No. marql is read-only by default. We pull data from your existing systems, compute the operating model, and show it back as briefings, alerts, and cockpits without changing source records.

Usually read-only API credentials, database views, or scheduled exports approved by the owner, accountant, or IT lead. We scope the least invasive path first.

That depends on the source. Some systems can sync hourly, others daily. The goal is operational visibility, not unnecessary polling, so we match the cadence to the workflow.

That is still workable. marql can normalize scheduled CSV or Excel exports as a first integration path, then move to API or direct connector paths later if needed.

If you want to validate more stores, roles, or metrics, we usually turn that into a 7-day pilot.

Yes. Mixed-system networks are a core use case. The normalization layer is specifically built to reconcile different POS, ERP, accounting, and export formats across the chain.

Ready to see your own data flow?

We will scope the least invasive connector path first.