ML Insights
Turn everyday business data into clearer next steps.
NexaPOS analyses your sales, customer, stock, supplier, and access patterns to surface explainable, advisory insights for better decisions.
Why teams trust ML Insights
- Built from your business data
- Explainable recommendations
- No customer data sent to external AI
- Your team stays in control
ML Insights is advisory, explainable, rule-based intelligence inside NexaPOS. It works from your own NexaPOS records. Customer and transaction data stays within NexaPOS and is not sent to external AI or LLM services.
ML Insights is included in the Multi-Branch plan and available as an add-on on every other plan. See plans and add-ons
Sales and demand forecasting
Plan ahead with practical sales forecasts.
See likely sales demand before the next trading period, so teams can prepare inventory, staffing, and purchasing with more confidence.
- Sales forecasts for the next 7, 14 and 30 days.
- Demand forecasts by product and by category.
- Branch-by-branch comparison for authorised teams.
- Stockout candidate alerts that flag products whose projected demand is close to the stock on hand.
Forecasts are estimates based on your sales history. They support planning; they do not guarantee results.
Customer intelligence
Know which customers need attention.
Spot customers whose buying pattern is changing, understand why they may be at risk, and build more relevant retention plans.
- High-value repeat customers.
- New customers.
- Loyalty-active customers.
- At-risk and lapsed customers.
- A churn-risk explanation for each customer, showing why they were flagged.
- Reachable audience previews that respect marketing consent and channel preferences.
Insights never change customer records.
Campaign intelligence
Create more relevant campaigns.
Turn customer insights into campaign-ready audiences while keeping consent, approvals, and staff controls at the centre.
- Recommended campaign audiences for SMS, WhatsApp and email.
- Suggested segments, including high-value, lapsed, category-interest and promotion audiences.
- A safe handoff into campaign creation in the CRM.
- Existing opt-in, channel-preference, permission and audit controls stay in place.
Recommendations never send a campaign automatically.
Inventory and purchasing intelligence
Reduce stock risk and purchase with more clarity.
Prioritise replenishment based on sales rate, stock position, lead time, open orders, and supplier performance.
- Reorder recommendations.
- Projected stockout dates.
- Safety-stock guidance.
- Supplier performance scorecards.
- Visibility of alternative suppliers.
- Draft purchase orders created for review.
Purchase orders are created as drafts for team review; they are never submitted automatically. Stock levels and prices are never changed by an insight.
Access-risk intelligence
Strengthen operational access controls.
Surface access risks before they become operational problems, with clear routes for authorised teams to review and resolve them.
- Missing or excess administrator access.
- Unassigned or invalid branch access.
- Dormant user accounts.
- Open sessions after an account is deactivated.
- Administrators without two-factor authentication.
- Temporary-password hygiene.
Access insights flag what to review. Authorised administrators decide what to change.
How it works
Insightful, explainable, and always under your control.
Explainable
Each insight includes its reasons and a recommended next step.
Advisory
Your team reviews every recommendation before acting on it.
Private
No customer or transaction data is sent to external AI or LLM services.
Controlled
Access is role-based and limited to the branches each person works in.
Honest
When there is not enough data, ML Insights says so instead of inventing a result.
Who sees what
The right insights for each role.
Tenant administrators
Full ML Insights access, including access-risk insights and views across branches.
Managers
Sales, customer, campaign, purchasing and supplier insights for their own branches.
Cashiers
No accessCashiers do not see ML Insights.
Access is controlled by role, module approval, tenant, and branch.
From the blog
Further reading
- How to spot at-risk customers before they stop buyingSimple signals that a regular customer is drifting away, how to group customers by buying pattern, and how to plan retention contact that respects consent.
- Machine learning in retail: what it can and cannot do for your shopA plain guide for shop owners: what data-driven insights are, where they help a small retail business, where they stop, and how to judge any tool's claims.
- Sales and demand forecasting for small retailers: a practical guideHow sales forecasts work, how to turn them into reorder points and safety stock, and how to use them sensibly in a shop with one to five branches.
Make your next business decision with more context.
Explore explainable insights for sales, customers, campaigns, stock, suppliers, and operational access.