Is next week's plan ready?
AGENT · PLANT DEMAND PLANNER
The production forecast reaches the MES with detail by line and item.
Plant Demand Planner produces the weekly production demand forecast for each line and each item in the customer's manufacturing catalogue. It combines confirmed orders, sales estimates, current stock, production capacity. It feeds the customer's MES and MRP for operational planning.
Plant Demand Planner at work.
Weekly plan ready. Confirmed orders: 847 units across 12 items. Sales forecast: +120 units on A3 (ongoing campaign). A3 stock: 3-day buffer (alert). Line 2 capacity: Thu-Fri slot available to absorb the volume. Plan updated in MES and MRP. Flagging A3 for review.
Good. I'll validate the A3 criticality and confirm the plan.
Plan confirmed. The event is in the audit registry.
Why it exists.
Manufacturing production planning combines many factors in parallel: confirmed customer orders, sales forecasts from the commercial team, current stock levels, production capacity by line, raw-material supply constraints. For mid-large companies with detailed catalogues, the manual pattern of the production planning team is demanding and prone to synchronisation errors between systems.
How it works each week.
Plant Demand Planner activates weekly. It reads confirmed orders and sales forecasts, current stock from the WMS, production capacity by line from the MES, raw-material supply constraints. It calculates the consistent plan and feeds the customer's MES and MRP.
The decision stays with the manager.
The production planning manager reviews, validates, authorises. For significant criticalities (potential stock-out, line over-saturation) the agent triggers an explicit notification. Plan validation and authorisation stay with the manager.
From production planning manager to plant operations director.
Production planning manager
The manager reclaims the time spent on manual data collection and synchronisation across systems. The weekly plan arrives already built — capacity concentrates on validation and criticality management, not on data collection.
Line L3 · SKU-441 7-day fcst
Week 23 8,400 units
Week 24 9,100 units
MES · MRP fed
Variance vs plan: +6%
Plant operations director
The director sees line utilisation optimised on real data. Stock-out risks and line over-saturation surface before the production week, not during it, and become manageable.
Vendor · Acme Forge due diligence
Credit info B+ · 36 months
ESG · CSRD reporting complete
Geopolitics monitored
Approval bundle to the buyer
Supply chain manager
The supply chain manager has structured visibility on the procurement pipeline state: raw-material constraints enter the plan before they become an operational block.
Dock BK-4 12 slots today
09:30 · carrier A high priority
10:15 · carrier B standard
11:00 · carrier C delayed · rescheduled
Average wait time: 18'
Weekly plan for 20 production lines.
Plant Demand Planner activates every Sunday evening.
For a manufacturing company with 20 production lines and a catalogue of 150 items, Plant Demand Planner activates every Sunday evening to produce the plan for the following week. The agent reads current stock from the WMS, confirmed orders from the ERP (847 units across 12 high-turnover items), updated forecasts from the Sales CRM, declared capacity by line from the MES.
A3 3-day buffer + Sales campaign: potential stock-out.
The agent identifies a criticality: item A3 has a 3-day stock buffer and the sales forecast flags an ongoing promotional campaign that could generate +120 units of demand not yet in confirmed orders. It proposes allocating Thursday and Friday on line 2 for the extra A3 volume, and updates the plan in the MES and MRP.
The manager validates, the agent records.
The manager reviews the proposal in the dedicated Slack channel and confirms. The agent records the validation in the runtime audit registry. The week's plan is in production Monday morning, already aligned across WMS, ERP, MES, and MRP.
Declarative rules, MES/MRP integration in delivery.
The Plant Demand Planner rules are declarative. The production planning and supply chain team defines in a readable format the forecast patterns (weight of confirmed orders vs sales forecasts, time horizon), the capacity constraints by line, the order priority rules, the stock criticality alert thresholds. The rules live in the customer's repository, versioned, validated at agent startup.
Integration with the customer's MES, MRP, and ERP is delivered via a dedicated adapter during the project by the Exelab team.
- Language
- TypeScript (Node.js)
- LLM model
- customer's choice: Anthropic, OpenAI, Mistral, open source models hosted internally, AWS Bedrock for a private model
- Built-in controls used
- pii-detector, tool-rate-limit
- Native delivery channels
- Slack, Telegram, WhatsApp, OpenAI-compatible HTTP
- MES (Manufacturing Execution System) + MRP + ERP integration
- dedicated adapter delivered during the project
- Memory
- persistent per instance, pgvector + PostgreSQL FTS on historical production patterns
- Registry
- append-only, queryable with a standard SQL client
Frequently asked questions about the agent.
Not without confirmation from the manager. The agent produces the plan and proposes it to the production planning manager. For significant criticalities (potential stock-out, line over-saturation) it triggers an explicit notification. Plan validation and authorisation stay with the manager.
Integration with MES, MRP, and ERP is delivered via a dedicated adapter during the project by the Exelab team. Technical feasibility depends on the customer's specific system. Exelab works with the main manufacturing systems (SAP, Oracle, proprietary systems).
Supply constraints (supplier lead times, open inbound orders, minimum safety stock levels) enter the declarative rules configured by the supply chain manager. The agent applies them in the plan calculation and flags cases where a supply constraint threatens line saturation.
The typical pattern for Plant Demand Planner is 14-20 weeks. Discovery 2-3 weeks, rule and forecast pattern configuration with the production planning team 4-5 weeks, MES/MRP/ERP integration 6-8 weeks, hand-off to the responsible team 2-3 weeks.
From a 30-minute conversation to the squad in production.
A 30-45 minute conversation to understand how Plant Demand Planner would configure to the customer's case. Which MES/MRP/ERP systems, which catalogue size, which production capacity constraints.