Amplinth Nexus is a graph-native supply-chain orchestration platform that plans, senses, watches and acts on a single Neo4j substrate. The same graph that plans your network watches it and helps act on every deviation — concurrent planning, ML demand intelligence and a graph-native control tower, with a natural-language copilot and governed autonomy.
So the plan is stale the moment a node deviates — and the cost shows up as expedites, stockouts and trapped working capital.
The disciplined Amplinth loop, run on a single Neo4j network model — so there is no plan-versus-reality gap to reconcile.
Ingests demand, supply, inventory, orders, product/BOM, network and cost data into one Neo4j knowledge graph; external signals feed demand sensing.
Graph algorithms (GDS) for network monitoring and risk propagation; ML forecasting for demand; concurrent planning over the shared model.
Constrained plans, scenario comparisons, reorder and rebalancing proposals, ranked network-aware exceptions, and NL answers from the copilot.
Control-tower alerts and Nexus Auto drive corrective actions with human-in-the-loop approval; multi-hop rebalancing moves stock across nodes.
Every operation is audit-logged — working capital released, service level protected, network cost avoided.
Nexus is a federation of supply-chain capability services sharing one Neo4j model, one tenant fabric and one audit spine. Deploy only the modules you need and scale them independently — each is its own FastAPI service with its own CI/CD.
From demand to supply to S&OP.
Concurrent demand, supply and capacity planning with scenario sandboxes for risk-free what-if — every change propagates instantly across the graph so you see the network-wide impact before you commit.
ML demand intelligence — a Prophet/XGBoost-style forecasting ensemble with demand sensing that adjusts to external signals. Coming soon causal-driver decomposition that explains why demand is moving.
A structured S&OP workflow with a dedicated demand-review step, so sales and operations align on one plan grounded in the same graph.
The execution core.
The deepest module: real-time stock, ABC/XYZ/FSN classification, reorder and norms, SLOB and aging analysis, FIFO/LIFO/WA valuation, financial analytics — and a multi-hop rebalancing engine that moves stock across nodes a relational tool can't model cleanly.
Production orders, purchase orders, master production schedule, MRP, supply analytics and collaborative demand planning — the supply side of the constrained plan.
Order management and fulfilment across the network — one of the largest router surfaces in the platform, covering the full order lifecycle.
Product master, BOM, configuration, pricing, categories, media and import/export — the product backbone of the network graph.
See and steer the live network.
Proactive monitoring: operational alerts, variance tracking, GDS-based network monitoring, an AI-powered analytics summary and a signals/pulse cockpit (Signal Inbox) that turns noise into ranked, network-aware exceptions.
Network performance — hub operations, transport and network costs, network risks, vehicle utilisation, partner performance and service-level tracking.
Cost-to-serve and cost analytics across the network, so every plan and rebalance can be weighed against its true landed cost.
The graph foundation — network and operations entities, the Neo4j network model, and the largest router surface that every other service builds on.
Ask, act, account.
Query the whole supply chain in plain English, grounded in your tenant graph — no dashboard hunting, just answers from the live network.
Autonomous exception resolution with human-in-the-loop. The core loop runs today; the library of governed autonomous actions is growing.
Carbon and Scope-3 emissions tracking for ESG — foundational today, with a fuller suite on the roadmap.
The module breadth — concurrent planning, the deep Inventory module, the Watch control tower, NetPerf and the NL copilot, all on a multi-tenant, audited platform — is deployed and running. The forecast-accuracy, service-level and savings figures below are targets or modelled values, held until a cleared client baseline exists.
The same Neo4j model that plans your network watches it and helps act on it — one substrate, not two products bolted together.
Coming soon, not claimed today: causal forecasting (explaining why demand moves) and a broader library of governed autonomous actions are near-term; a full Scope-3 / ESG suite and a certified ERP/TMS/WMS connector catalogue are on the roadmap. We label maturity honestly so you can plan with confidence.
Connect your demand, supply, inventory, order and network data into one Neo4j graph. Compose the platform from only the modules you need — each deploys and scales independently with its own CI/CD. Operate the connected loop in production, where the same graph plans, senses, watches and acts, and every operation lands in an immutable audit trail.
The same Neo4j network model that plans the chain also watches and acts on it. Planning tools and control towers are usually separate products with separate data — Nexus closes that gap on a single substrate.
Multi-hop inventory rebalancing, GDS-based network monitoring and risk propagation are first-class capabilities — not features bolted onto a relational core that fights the network shape.
Forecasting built for planner trust, a natural-language copilot over the graph, and autonomous actions that stay human-in-the-loop with an immutable audit trail — across 14 independently deployable services.
Book a 30-minute demo on your toughest planning-to-execution problem — we'll show you the Inventory module, the Watch control tower and the copilot, all on one graph.