Locus Dispatch Agent
AI agent · Workflow / back-office agent · Operations & supply chain by Locus
Automated, constraint-based dispatch for enterprise logistics excellence.
Locus Dispatch Agent is a specialized AI agent within the Digital Supply Chain Officer (DiSCO) framework, designed to automate transportation and logistics decisions for enterprise retailers, 3PLs, and carriers. The agent operates against 250+ real-world operational constraints, including time windows, vehicle capacity, and driver skills, to ensure feasible execution. It supports configurable autonomy levels from L1 recommendations to L3 full autonomous execution, allowing human planners to maintain override control. Locus Dispatch Agent integrates via API with existing enterprise systems like
Agent specs
The full spec card (context, pricing per token, modalities and more) on AntSpace. Sign up free to see the full spec card.
Category: Transportation Management
- Integrations
- SAP, Oracle, Salesforce, TMS, OMS, WMS, ERP, CRM systems
- Operational constraints
- 250+
- Decision speed
- ~2 minutes
- Autonomy levels
- L1 to L3
- Integration type
- API-first
- Supported fleets
- Owned, contracted, spot, and gig
- Automates dispatch across 250+ real-world constraints
- Supports autonomy levels from L1 (recommendations) to L3 (autonomous)
- Generates optimal plans in approximately 2 minutes
- Powers 99.5% on-time delivery rates for enterprise clients
- Governed by explainability and human-in-the-loop mechanisms
Overview
Locus Dispatch Agent is a specialized AI agent within the Digital Supply Chain Officer (DiSCO) framework, designed to automate transportation and logistics decisions for enterprise retailers, 3PLs, and carriers. The agent operates against 250+ real-world operational constraints, including time windows, vehicle capacity, and driver skills, to ensure feasible execution. It supports configurable autonomy levels from L1 recommendations to L3 full autonomous execution, allowing human planners to maintain override control. Locus Dispatch Agent integrates via API with existing enterprise systems like SAP, Oracle, and Salesforce to act as a system of execution above the system of record.
Dispatch and Allocation
- Constraint-Based Decisions: Every decision is bound by 250+ real-world operational constraints, including time windows, distance, vehicle types, and driver skills.
Governance and Control
- Configurable Autonomy Levels: Operators can set autonomy levels from L1 (recommendations) to L3 (full autonomous execution) based on their operational needs.
- Governance Mechanisms: The agent is governed through mechanisms including Explainability, Traceability, and Evaluation to ensure enterprise trust.
- Human-in-the-Loop Control: A Human-in-the-Loop mechanism allows teams to override any recommendation or decision made by the agent.
- Execution Sandbox: The agent operates within an Execution Sandbox to validate decisions before they are carried out on the ground.
Real-Time Execution
- Continuous Re-optimization: The agent continuously monitors execution and re-optimizes the remaining plan while the shift is still running to account for delays or exceptions.
What it does
- Automated job allocation
- Daily scheduling for pickups and deliveries
- Real-time route re-optimization
- Dynamic order injection and sequencing
Published results
- 99.5% On-time delivery rates
- Up to 20% reduction in logistics costs
- 81% Time back in planning efforts
Use cases
- Retailers re-sequence active delivery routes in real-time when customer availability changes or traffic delays occur.
- Field service operators match urgent service calls to technicians based on certification requirements and geographic proximity.
Specifications
- Channels
- API, in-app (Driver Companion App), website (Control Tower dashboard)
- Integrations
- SAP, Oracle, Salesforce, TMS, OMS, WMS, ERP, CRM systems
- Data sources
- Traffic data, weather feeds, GPS and telematics, mapping providers, carrier tracking APIs, fuel pricing, regulatory data
- Human control
- Autonomy levels L1 to L3; Human-in-the-loop override; Execution sandbox for validation; Explainability for decision logic
- Deployment
- Cloud-based platform; Global presence in United States, Europe, Asia Pacific, Middle East & Africa
- Operational constraints
- 250+
- Decision speed
- ~2 minutes
- Autonomy levels
- L1 to L3
- Governance mechanisms
- 6 (Explainability, Traceability, Evaluation, Autonomy, Sandbox, Human-in-the-Loop)
- Integration type
- API-first
- Supported fleets
- Owned, contracted, spot, and gig
- On-time delivery rate
- 99.5%
How it works
- Stand Up and Deploy: Forward Deployed Engineers embed with the customer's team to stand up the platform and configure custom agents and workflows.
- Tune and Graduate: The system is tuned against real-world data and operational rules, graduating from recommendations to higher autonomy levels.
- Autonomous Operation: The customer's team takes ownership of the agentic operations, with the AI agent running order-to-settlement autonomously.
FAQ
What is the Locus Dispatch Agent?
The Locus Dispatch Agent is a specialized AI agent within the DiSCO framework that automates dispatch allocation, matching orders to drivers based on 250+ real-world constraints.
Can I control the level of autonomy for the agent?
Locus offers three autonomy levels: L1 for recommendations, L2 for partial autonomy, and L3 for full autonomous execution, all governed by human-in-the-loop mechanisms.
Does the agent handle real-time changes during a shift?
Yes, the agent continuously monitors live execution signals like traffic and driver updates to automatically re-optimize routes when delays or exceptions occur.
Added by AntSpace from public sources (locus.sh).