Each agent handles one part of the MRO problem, reads live data to do it, and writes its result back to your ERP only after a human approves.
Here is what each one does.
1
Module 1 of 9
Spare Part Search
Search AI
An agent that instantly surfaces full part details, criticality, lead time, safety levels, and technical specs, plus an AI chat for plain-text questions and on-demand reports. No manual cross-checking across ERP screens or spreadsheets.
Schlüssel-Fähigkeiten
- Search any part by Material Code, MFR, MPN, or category in one query.
- Built-in AI chat: ask any question in plain text or pull a report on the go.
- Multi-plant, multi-site visibility from a single search, with no system switching.
Business Value
- Planners find parts in seconds, not by toggling ERP screens.
- AI chat cuts report turnaround from hours to minutes.
- One query gives full network visibility.
The agentic edge: standard ERP search needs exact codes and screen-hopping. This agent reads plain-language intent, retrieves across plants, and compiles the full answer itself.
2
Module 2 of 9
Kritikalität der Ausrüstung
Asset Intelligence
An agent that scores equipment criticality from maintenance history, safety and production impact, MTBR, MTBF, and linked spares, enriched with notification and RCA insight.
Scoring Parameters
- Maintenance history plus MTBR and MTBF combined into one composite score.
- Safety and production impact, weighing linked spares, functional location, and redundancy.
- Notification and RCA mining surfaces failure patterns manual reviews miss.
- Greenfield baseline scores new equipment instantly, with no coverage gaps.
Business Value
- Replaces weeks or months of manual criticality work with on-demand scoring.
- New equipment gets a reliable baseline from day one.
- Approved ratings write back to ERP/CMMS, keeping the record current.
The agentic edge: standard criticality is a one-time consultant study in spreadsheets that decays. This agent re-scores continuously from live history, RCA, and notification mining, re-ranking in hours, not a re-engagement project.
3
Module 3 of 9
Ersatzteile Kritikalität
Inventar Intelligenz
An agent that scores spare part criticality from consumption history, associated equipment, lead times, and vendor count, enriched with notification and work order insight for a transparent, component-level breakdown.
Schlüssel-Fähigkeiten
- Aggregates consumption, equipment installs, and quantities across all plants into one score.
- Weighs lead time and vendor count, so single-source parts score higher.
- Mines notifications and work orders for context, improving accuracy over time.
- Writes approved ratings back to ERP/CMMS with no manual re-entry.
Business Value
- Single-source, long-lead parts flagged automatically.
- Component-level scores replace gut-feel stocking with data-backed policy.
- Scores improve as new work orders and notifications feed signal.
The agentic edge: standard tools rank on consumption alone (ABC/XYZ). This agent fuses supply risk, equipment linkage, and mined work-order context into one explainable score, so single-source critical parts cannot hide in a C-class bucket.
4
Module 4 of 9
Obsolescence Check + Substitutes
Obso Check AI · Spare Seek AI
Two agents working in tandem. Obso Check AI flags parts discontinued by the manufacturer, while Spare Seek AI recommends substitutes plus the vendors that supply them and at what cost.
Schlüssel-Fähigkeiten
- Shows the full discontinuation spectrum, from fully discontinued to available through select vendors.
- Surfaces long-discontinued stock sitting quietly in inventory, freeing working capital.
- Spare Seek AI recommends substitutes with vendor names and landed cost.
- Writes approved status and substitute updates back to ERP/CMMS.
Business Value
- Idle obsolete stock cleared before it becomes a write-off surprise.
- Planners get qualified substitutes with vendor and cost, not just a flag.
- Prevents stoppages caused by ordering a part that is no longer made.
The agentic edge: standard tools flag obsolescence at best. These tandem agents detect it and source qualified substitutes with vendor and landed cost, so planners get a ready-to-act fix, not a research task.
5
Module 5 of 9
Aktienumsatz
Fast / Slow / Non-Moving
An agent that continuously classifies every part as fast-moving, slow-moving, or non-moving, then recommends removing dead stock to free working capital and right-size inventory across every plant.
Schlüssel-Fähigkeiten
- Classifies every part as fast, slow, or non-moving, per plant, continuously.
- Delivers a stock-velocity report across all critical items.
- Recommends write-off or transfer to a plant where the part is more useful.
- Feeds directly into the reorder point and demand forecasting agents.
Business Value
- Dead stock surfaced network-wide, not just at one plant.
- Velocity data feeds reorder and forecasting automatically.
- Reduces carrying costs without manual audits or spreadsheet reviews.
The agentic edge: continuous, network-wide classification plus a recommended action, so dead stock becomes redeployable capital, not a static report.
6
Module 6 of 9
Nachfrageprognose
Multi-Variant AI Forecasting
A multi-variant agent that monitors consumption history, equipment failure history, and other parameters, capturing seasonality, spikes, and anomalies to forecast spare parts demand up to two years out.
Wie es funktioniert
- Aggregates consumption, failures, and operational data into one signal set.
- Multi-variant models detect seasonality and anomalies.
- Segments demand into preventive, corrective, production, and shutdown.
- Publishes a rolling two-year forecast by project, plant, and part.
Business Value
- Cuts emergency buys and excess safety stock at the same time.
- Shutdown and seasonal demand modelled separately, not lumped into an average.
- A two-year runway gives procurement room to negotiate lead times.
The agentic edge: standard forecasting is single-variable moving averages. This agent learns failures, seasonality, shutdowns, and anomalies by demand cause, with a two-year runway standard tools cannot see.
7
Module 7 of 9
Bestellpunkt
Advanced Inventory Modelling
An agent that dynamically calculates min and max levels, safety stock, and reorder dates and quantities for every material in the plant, tuned to actual supplier lead-time variability and each part's criticality tier.
Schlüssel-Fähigkeiten
- Sets min/max, safety stock, and reorder date and quantity per part.
- Builds actual supplier lead-time variability into every calculation.
- Runs on demand or on a schedule, based on your preference.
- Raises proactive alerts before a critical material runs short.
- Writes approved settings back to ERP/CMMS after approval.
Business Value
- Safety stock reflects real supplier behavior, not guesswork.
- Shortage alerts give planners time to act before a work order is at risk.
- Min and max always reflect the latest calculation.
The agentic edge: standard min/max is set once and hard-coded until the next cleanup project. This agent recalculates dynamically from live lead-time variability and criticality tier, so shortages are prevented before they happen, not expedited after.
8
Module 8 of 9
Work Order Planning & Supplier Reliability
Vendor Intelligence
A proactive warning system that flags upcoming work orders at risk from insufficient material or questionable vendor reliability, delivering a prioritised task list with recommended order dates and quantities.
Schlüssel-Fähigkeiten
- Checks on-hand stock and upcoming POs against planned work order demand.
- Continuously scores supplier performance and flags low-reliability vendors on critical parts.
- Prioritises long-lead-time items earliest, where late ordering hurts most.
- Writes approved order dates and quantities back to ERP/CMMS.
Business Value
- Planners see on-hand, incoming POs, and live demand before a work order falls short.
- Vendor reliability scoring catches risky suppliers before they delay a job.
- A risk-ranked task list replaces firefighting.
The agentic edge: standard MRP checks on-hand stock only when the order drops. This agent looks ahead at POs, demand, and vendor reliability before the order is at risk.
9
Module 9 of 9
Interplant Transfer Intelligence
Unified Multi-Plant Visibility
An agent that identifies inter-plant transfer opportunities from material need, open POs and requisitions, and dead stock or excess inventory across the network, sourcing internally before triggering an external purchase order.
Schlüssel-Fähigkeiten
- Spots transfer opportunities from material need, open POs/PRs, and dead stock across all sites.
- Surfaces excess inventory idle at one plant that another plant urgently needs.
- Sources long-lead-time materials from nearby plants, cutting shutdown risk.
- Writes approved transfer orders back to ERP/CMMS after approval.
Business Value
- Treats the whole network as one warehouse: excess at Plant A fills a gap at Plant B.
- Internal sourcing cuts emergency external POs and their premium cost.
- Dead stock cleared through transfers reduces write-offs network-wide.
The agentic edge: standard ERPs treat each plant as a silo. This agent matches one plant's need to another's excess before any PO is raised, so the network becomes one virtual warehouse.
