ActiveAgents™ are the AI agents that run inside Manhattan Active solutions, working on the same live operational data and under the same rules as your people. In the GCC and Africa, Supply Chain Junction implements and supports the Manhattan Active solutions they run in.
What does “built in, not bolted on” mean?
Most enterprises are adding AI assistants, copilots and chatbots on top of systems that are already fragmented, and AI bolted onto fragmentation only amplifies it. Manhattan’s answer is that AI is built into Manhattan Active: its agents live inside the system of execution, aligned to how supply chain teams actually work.
Video: Manhattan Associates. Turn on captions with the CC button.
| Generic AI | ActiveAgents |
|---|
| Data | Reasons over stale, disconnected data | Reason over one real-time source of operational truth |
| Actions | Suggests actions it can’t safely take | Governed by the same rules as your operators |
| Knowledge | Knows operations in general | Know your operations, and get smarter every 90 days |
Source: Manhattan Associates.
Which agents does Manhattan offer?
Manhattan lists ready-to-deploy agents across its Manhattan Active solutions. Examples it names:
- ActiveWarehouse™: the Labor Agent, which reports on employee performance and utilisation and helps reallocate spare labour; the Wave Coordinator Agent, which investigates wave issues such as short inventory; and the Warehouse Associate Agent. Manhattan also lists Cycle Count Planning, Create ASN, Verify ASN and Shipping Supervisor agents.
- ActiveTransportation: Transportation Planner, Dispatcher, Appointment Management, Rate Management and Freight Payment agents.
- ActiveOrder: Contact Center, Order Creation, Backorder Monitoring and Late Delivery Refund agents.
Manhattan adds new agents and skills every 90 days, so check manh.com for the current catalogue.
Can you build your own agents?
Yes. Manhattan Agent Foundry™ lets business users describe what they need in plain English, start from a template, or have developers build agents from scratch. Agents built in it follow the guardrails administrators set, and it supports the MCP and A2A standards for working with other agents.
What are Manhattan’s customers reporting?
Manhattan reports these results from its own customers. It does not name them, and none are Supply Chain Junction clients.
- A global industrial manufacturer cut dock cycle time by up to 27% and unlocked nearly 15% more dock capacity.
- A 200-store grocery and pharmacy retailer moved from base agents to custom agents in 30 days and cut cycle times by about 30 minutes.
- A global healthcare distributor improved cycle time by 25% and cut shortages by more than half.
- A 700-store pet retailer is unifying its supply chain and, Manhattan says, compounding value with every release.
Source: Manhattan Associates. Results are Manhattan’s figures for its customers and are not a promise of the same outcome.
How do you get ready for ActiveAgents?
Agents act on live data in Manhattan Active, so they are only as good as the foundation under them. That foundation is what a Manhattan implementation builds, and it is where we work with you:
- Agreed business processes. Agents follow the same rules as your operators, so the rules have to be agreed and written down first.
- Warehouse and master data readiness. One real-time source of truth depends on clean data.
- Your team structure. Who owns each process, and who decides how far an agent may act on its own.
- Training and change management. Your people need to know what the agents do and when to step in.
A first site typically takes 6–12 months, from design start to stable operation. See how an implementation runs or read what drives timeline.
Frequently asked questions
Who implements AI-powered warehouse systems in the Middle East?
Supply Chain Junction is Manhattan Associates' GeoPartner for Africa and the Middle East, with over 90 implementations in the GCC and Africa and teams in South Africa, Saudi Arabia and the United Arab Emirates. We implement and support the Manhattan Active solutions that Manhattan’s ActiveAgents run inside.
What is the difference between an AI agent and a chatbot?
Manhattan describes a chatbot as mainly following scripts or decision trees for simple tasks, while an agent is goal-driven and can take multi-step actions across systems.
How are ActiveAgents kept safe?
Manhattan says its agents are governed by the same rules as your operators, are grounded in Manhattan Active data to reduce hallucinations, and have every decision logged and explained. Administrators set the guardrails.
Can we build our own agents?
Yes. Manhattan Agent Foundry™ lets you describe what you need in plain English, start from a template, or have developers build from scratch. It supports the MCP and A2A standards.
What do ActiveAgents depend on?
They act on live data in Manhattan Active, so they depend on the processes, warehouse data and master data behind it. Getting those right is the design phase of an implementation.
How long does an implementation take?
A first site typically takes 6–12 months, from design start to stable operation.
Where the agents work
GL
SA
UAE
ZA
Explore
