PRODUCTS
DEMAND PLANNINGOS
Demand PlanningOS
BY Nehlum
AI SUPPLY CHAIN PLANNER
Demand PlanningOS: AI Demand and Supply Planning Software
Demand PlanningOS is the AI planner for your supply chain. When an order arrives, it tells you what to buy, how much, from whom and by when. It watches every supplier and shipment, warns you before a delay reaches production, and learns from every delivery. It sits on top of your ERP, warehouse, procurement and fleet systems. It plans and warns. You approve.
- Connects ERP, WarehouseOS, ProcureOS, FleetOS & shipping data
- Reads the order
- Finds the shortage
- Counts back from delivery
- Picks the supplier on evidence
- Warns before delays
- You approve
01THE QUESTIONS
Be honest — how many of these did you ask this week?
When does the supplier say it will deliver, and when does it really deliver?
How often does a late material put a production batch at risk?
What does one lost production day cost you?
How much money is sleeping in your warehouse?
What did your last order really cost to deliver?
Most companies only see the purchase price. Demand PlanningOS shows the rest.
SAR at risk every month
before anyone notices.
in production-day exposure in idle stock ()
Illustrative estimate. Your figures will differ.
How we calculate
- Planning hours = orders per month × hours per order
- Planning cost = planning hours × hourly cost
- Delay exposure = orders per month × share hit by a late material × days late × cost of a stopped day
- Idle stock cost = slow-moving stock value × carrying rate ÷ 12
- At risk per month = delay exposure + idle stock cost + planning cost
02THE BROKEN CHAIN
Twelve links between an order and a delivery. None of them talk to each other.
Sales knows the order. The warehouse knows the stock. Procurement knows the supplier. Logistics knows the ship. Nobody sees the whole chain, so problems are found on the day they hurt.
Before 12 hand-offs · problems found on the dayAfter 1 loop · problems found weeks ahead Illustrative
Select a link to connect it.
Links connected: 12 / 12
Every delivery makes the next plan smarter.
What if every order told you, the day it arrived, exactly what to buy, from whom, and by when?
What if you knew a ship was late before your production line did?
03THE FIX
Meet Demand PlanningOS. It does the planning. You make the call.
-
01What do we need to buy?
Today A spreadsheet and a phone call to the warehouse.AI actionAI explodes the recipe and nets stock, reservations and inbound goods.
Human outcome The exact shortage, per delivery. -
02When must we order?
Today Planned on the lead time the supplier promised.AI actionAI counts back from the delivery date using the actual lead time and clearance days.
Human outcome Order dates that hold. -
03Which supplier will deliver?
Today The cheapest quote wins.AI actionAI rates suppliers on real deliveries and picks the lowest risk-adjusted cost.
Human outcome Supplier choice on evidence. -
04Which ship?
Today The cheapest freight is booked.AI actionAI compares sailings on freight, transit, reliability and safety stock.
Human outcome The right vessel, not the cheapest. -
05The ship is late. What breaks?
Today Found out at the port.AI actionAI recalculates the whole plan and ranks the fixes.
Human outcome Act weeks before the line stops. -
06Why is money stuck in stock?
Today Nobody prices the days stock sits.AI actionAI prices every day stock sits and how chemicals lose strength.
Human outcome Capital released, waste visible. -
07Where did the material go?
Today Losses are guessed at month end.AI actionAI compares issued vs output vs recipe, by batch, line, shift and lot.
Human outcome Losses traced, not guessed. -
08What did the order really cost?
Today Only the purchase price is known.AI actionAI adds up the true cost to serve from purchase to proof of delivery.
Human outcome Real margin per order and customer. -
09Who approves?
Today Making the callAI actionNo AI action. This step belongs to a person.
Human outcome You do.
Swipe or select a card to move on.
Every number comes from a transparent calculation you can click back to its source. The AI explains and drafts. Every purchase, booking and transfer waits for your approval.
04DATA IN, DECISIONS OUT
Everything it knows, it gets from your systems. Everything it gives back, you can act on.
Demand PlanningOS connects to the systems you already run. It reads orders, stock, purchases, shipments, production and deliveries, turns them into one connected picture, and sends back decisions, alerts and drafts for your approval.
Outputs available: 13 / 13 (all full)
Start in a week with what you have. Connect the rest when you're ready.
Diagram: seventeen data sources feed Demand PlanningOS, which connects, calculates and explains, and returns thirteen outputs. Outputs that write back to another system pass through your approval first. The same content is available as a table below.
Data in
Nehlum OS family
Enterprise systems
Communication
Logistics data
Market & finance
Demand PlanningOS
Approval gate
- ConnectIdentity map: one name for every material, supplier and customer.
- CalculateShortage, timing, supplier score, cost, risk. Code, not guesses.
- Explain & recommendPlain-language reasons and drafts, waiting for approval.
Decisions out
Select a source or an output to see what depends on what. Switch a source off to see what stops.
Latest eventVESSEL_ETA_CHANGED → plan recalculated, 3 fixes ranked
Show as table
| Group | Integration | What it reads | What it sends back | Connection |
|---|---|---|---|---|
| Nehlum OS family | WarehouseOS | Stock on hand, reserved and available; lots, batches and expiry; ASNs; goods receipts (ordered, received, damaged, rejected, returned); recipes and bills of materials; material issued to production; finished-goods receipts; days on hand. | Put-away and stock-transfer requests, on approval. | API |
| Nehlum OS family | ProcureOS | Supplier master; RFQs, quotes and comparisons; purchase orders; approval status; RFQ win and loss history. | Procurement requirements and RFQ drafts, on approval. | API |
| Nehlum OS family | FleetOS | Vehicles, drivers, compliance (Istimara, licence), trip costs, on-time and damage history, proof of delivery. | Trip requests for each delivery, on approval. | API |
| Nehlum OS family | InquiryOS | Customer orders and production plans arriving by email, already extracted into structured records. | — | API |
| Enterprise systems | ERP: SAP S/4HANA · Oracle · Microsoft Dynamics 365 | Sales orders and delivery schedules, item master, production plans, standard costs, finance postings. | Approved purchase orders and requirements, if the client allows write-back. | API · Excel / CSV |
| Enterprise systems | CRM | Sales pipeline, customer forecasts, contract volumes. | — | API · Excel / CSV |
| Enterprise systems | E-commerce / online orders | Online order lines and demand by SKU. | — | API |
| Enterprise systems | Quality / lab records | Certificates of Analysis, tested concentration at receipt and at issue, QC pass or fail. | — | API · Excel / CSV · Manual |
| Communication | Email mailbox | Customer orders, production plans, supplier quotes and replies, questionnaire answers. | Approved RFQs, chasers, award and decline notices. | Email agent |
| Communication | Historical data import | 24 months of sales, purchases, receipts and production, in the client's own format. | — | Excel / CSV |
| Logistics data | Ocean schedules (Linescape) | Sailings: vessel, voyage, ports of loading and discharge, departure and arrival dates, transit, transshipment. | Draft booking requests, on approval. | Browser agent |
| Logistics data | Vessel tracking (MarineTraffic / carrier APIs) | Live vessel position and ETA changes, by bill of lading. | — | Browser agent · API |
| Logistics data | Air cargo tracking | Air waybill milestones. | — | API |
| Logistics data | Clearance days per lane | Typical customs clearance time per import lane, or from CustomsOS. | — | Manual |
| Market & finance | Exchange rates | Daily rates; USD/SAR peg 3.75; EUR and others. | — | API |
| Market & finance | Calendar & seasonality | Ramadan, Hajj, summer peaks, public holidays. | — | Built-in |
| Market & finance | External signals (optional) | Market and economic indicators. | — | API |
| Output | Who uses it | Goes to |
|---|---|---|
| Demand forecast | Demand planner | Dashboard · drives purchase and production |
| Exact shortage per delivery | Planner | Dashboard |
| Order-by dates | Buyer | Dashboard · ProcureOS |
| Procurement requirements & RFQ drafts | Buyer | ProcureOS, on approval |
| Supplier ratings with evidence | Procurement head | Dashboard · supplier feedback |
| Shipping recommendation & draft booking | Logistics | Forwarder or carrier, on approval |
| Delay alerts with ranked fixes | Planner, operations | Decision queue · email |
| Inventory actions & carrying cost | Operations, finance | Dashboard · WarehouseOS, on approval |
| Production yield & loss alerts | Production manager | Dashboard |
| Distribution plan & trip drafts | Operations | FleetOS, on approval |
| True cost to serve | Executive, finance | Dashboard · reports |
| Learning updates | Everyone | Next plan |
| Approval & audit trail | Management, audit | Audit log |
It does not replace these systems. It reads from them, connects them through one shared identity map, so that “MEG”, “RM001”, “MEG-01” and “Mono Ethylene Glycol” are recognised as the same material, and writes back only approved actions.
Demand PlanningOS reads from your systems and writes back only what you approve. Every number it shows can be traced to the record it came from.
05SEE IT WORK
Give Demand PlanningOS a customer order. Watch it plan, warn and learn.
Pick an order and press Run. Demand PlanningOS reads it, finds the shortage, picks the supplier and ship, and then a ship runs late. The fix is yours to approve.
Simulation with sample data- Read
- Need
- When
- Who
- Ship
- Alert
- Fix
- Your approval
- Learn
1 · Pick an order
Select one, or drag it into Demand PlanningOS.
2 · Demand PlanningOS
Drop an order here
SO-2026-1184
- Customer
- Gulf Coolants Co.
- Product
- Coolant 50/50
- Deliveries
- 6 monthly deliveries of 1,000 t, first on 15 Nov
- Delivery sitePlease confirm 84%
- Dammam plant
- Payment termsPlease confirm 76%
- 60 days
Order captured in 40 s
Days of warning18
Typical: found on arrival day · Demand PlanningOS: 18 days earlier Illustrative
What to buy
Fed byWarehouseOSERP| Material | Needed | Available | Short |
|---|---|---|---|
| MEG | 1,345.6 t | 470 t | 875.6 t |
| Inhibitor | 85.9 t | 52 t | 33.9 t |
| Dye | 1,431 kg | 1,400 kg | 31 kg |
| IBC totes | 2,679 | 1,540 | 1,139 |
| Water | 1,430 t | Made in house |
Needed for the first three deliveries
When to order
Fed byProcureOS historyClearance days- Order by 30 Oct
- Production starts 8 Dec
- Delivery 15 Dec
Counted back on the actual lead time of 36 days, not the promised 35, plus 3 safety days.
Who to buy from
Fed byProcureOSWarehouseOS receiptsExchange rates
Supplier A
82
Promised 30 d
Actual 34 d
Medium risk
Supplier B
97
Promised 35 d
Actual 36 d
Low risk
Supplier C
64
Promised 28 d
Actual 41 d
High risk
Supplier X is rated better but needs 44 days and would miss the date by 8 days. Supplier Y can make it. Supplier switched
Requirement sent to ProcureOS
Which ship · Rotterdam → Dammam
Fed byLinescape- Vessel 1 Freight SAR 182,000 · 24 d transit Total cost SAR 214,000 Lowest total cost
- Vessel 2 Freight SAR 168,000 · 31 d transit Total cost SAR 236,500 Lowest total cost
- Vessel 3 Freight SAR 175,000 · 37 d transit Not feasible: arrives 6 d late Lowest total cost
Delay alert
Fed byMarineTraffic
Vessel from Jebel Ali: ETA +20 d
- Material
- Batch
- Delivery
- Customer
121 t of MEG short for the 15 Nov delivery · production at risk from 8 Nov · exposure SAR 255,000 At riskOn track
Ranked fixes
Fed byWarehouseOSProcureOS- 1 Buy 121 t from a local supplier who can deliver in time Cost SAR 9,300Exposure avoided SAR 255,000Net benefit SAR 245,700 Approved by you
- 2 Ask the supplier to truck 121 t by road from Jebel Ali Cost SAR 51,000Exposure avoided SAR 255,000Net benefit SAR 204,000 Approved by you
- 3 Transfer 60 t from the Jubail warehouse Cost SAR 5,300Exposure avoided SAR 126,400Net benefit SAR 121,100 Approved by you
- Air freight: not possible for bulk liquid
- Substitute material: none defined in the recipe
Your call.
Reason recorded — Demand PlanningOS learns from it.
What it learned
Fed byWarehouseOS GRNFleetOS proof of deliveryApprove a fix above. Demand PlanningOS then records what this order taught it.
Every delivery makes the next plan smarter.
Order captured · 4 shortages found · 7 orders timed · 1 supplier switched · delay caught 18 days early · SAR 255,000 exposure handled · Approved by you.
06OUTCOMES
Know before it breaks. Decide with the full picture.
Days of warning · Illustrative
Problem discovered
Problem discovered: 18 days earlierTime to fix it calmly
True cost to serve, per tonne · Illustrative
- 2,310Purchase price
- 148Freight
- 42Clearance
- 96Storage
- 121Production
- 71Losses
- 52Trucking
- 18Damage
- SAR 2,858True cost per tonne
Planned SAR 2,700Actual SAR 2,858
What did this product really cost to deliver to this customer? Now you know.
- Day 1shortage known when the order arrives
- 18 daysearlier warning on a late ship Illustrative
- 0purchases without your approval
- Everydelivery makes the next plan smarter
- Orders read and checked automatically
- Exact shortages per delivery date
- Order dates based on real supplier lead times
- Suppliers ranked on facts, not promises
- Ships chosen on total cost and risk
- Delays caught and fixes ranked before production stops
- Idle stock and material losses priced
- Production losses traced to batch, line and lot
- True cost to serve per order and customer
- Every approval and override recorded and learned from
See Demand PlanningOS on your own ordersRequest a demo
07FAQ
Questions planners ask before they start.
What does Demand PlanningOS actually do?
It's an AI supply chain planner. When a customer order arrives, it works out what to buy, how much, from whom and by when. It then watches suppliers and shipments, warns before delays hit production, and learns from every delivery.
Does it replace our ERP or warehouse system?
No. It sits on top of what you already use, such as SAP, Oracle, Microsoft Dynamics, WarehouseOS, ProcureOS and FleetOS, and connects them.
Which systems does it connect to?
Nehlum's WarehouseOS, ProcureOS, FleetOS and InquiryOS; your ERP (SAP, Oracle, Microsoft Dynamics 365), CRM and online orders; quality and lab records; your email; ocean schedules and vessel tracking; air cargo tracking; and exchange rates. You can start with Excel and email and add direct connections later.
Does the AI buy or book anything on its own?
No. Demand PlanningOS recommends and drafts. Purchases, supplier awards, vessel bookings, stock transfers and trips all wait for an authorised person to approve them, and every decision is recorded.
Can we trust the numbers?
Yes. Lead times, shortages, costs and supplier ratings come from transparent calculations, not AI guesses. Click any number to see the formula, inputs and source. The AI reads documents and explains results.
What data do we need to start?
Your sales, purchase and receipt history in Excel, plus an email inbox for orders and supplier replies. Direct system integration can follow when you're ready. Historical data is used as you provide it.
How are suppliers rated?
On what they actually deliver: on time or late, full or partial, damaged, rejected, returned, response speed and documents. Recent deliveries count more. Suppliers with too little history are marked as provisional.
Where is our data hosted?
Hosting and the AI model provider are agreed with each client before deployment. Tell us your data-residency requirements and we will confirm the options with you.
It plans, warns and recommends. You approve.
Plan smarter.
See it before it breaks.
Works withProcureOSWarehouseOSFleetOSCustomsOS