Skip to main content
Kardinal’s Route Optimization API turns your fleet, orders, and business constraints into an optimized route plan, and keeps it optimized as reality changes — a late delivery, a cancelled order, an urgent stop — without restarting from scratch. Most routing engines have to restart optimization from scratch after every change; Kardinal treats an update as a continuation of the same problem, not a new one. You control the trade-offs explicitly (fewer vehicles vs. faster service vs. lower cost) instead of treating the engine as a black box. It models operations across a wide range of logistics verticals, from last-mile delivery to waste collection and field-service dispatch — see the modeling examples by industry for how these constraints play out per vertical. This documentation covers technical integration and the concepts behind the optimization engine. It’s written agent-first: clear structure, explicit terminology, and worked examples an integration agent can act on directly. A few examples of what Kardinal models, by industry (not an exhaustive list — the same constraint model applies to other verticals too):
  • Parcel & express — a mixed van/bike/walker fleet, capacity-gated pickup/drop-off point (PUDO) rounds, and zone preference in a dense urban network.
  • Waste and bulk-collection — alternative disposal outlets, material changeover time, incompatible loads, and heavy-vehicle routing on a mixed fleet.
  • Cold-chain last-mile — single-mode-at-a-time refrigerated compartments, trusted-keychain deliveries, zone preference, and predictive traffic.
  • Field services — technicians dispatched from home, skill-based assignment, and opportunistic scheduling for lower-priority jobs.
  • Bulky & heavy goods — capacity-gated customer returns, a subcontractor revenue floor, equipment-dependent service time, and two-person crews.
  • Heavy long distance — reserved warehouse slots, load-aware CO2 accounting, self-closing consolidation loops, and a real overnight rest.
  • Retail last-mile — key-holder-only access windows, blended subcontractor/in-house costs, and pre-opening store drops.
See Modeling examples by industry for the full set.

Getting started

Send your first API call and get an optimized route in under 15 minutes.

Concepts

Understand how the optimization engine calculates and prioritizes solutions.

Guides

Authentication, advanced constraint modeling, real-time re-optimization.

API Reference

Endpoints, data model, error codes, and limits and quotas.

Modeling examples

See combined, realistic examples of the constraints that matter most for your industry.

Glossary

Unified terminology used throughout this documentation.

Security & compliance

EU hosting, encryption, GDPR, and certification status.

Pricing & SLA

Credit-based pricing, buying packs, and support/SLA commitments.
Kardinal uses a credit-based pricing model: every new account starts with 1,000 free credits, and additional packs are available directly from the Console — no sales conversation required to get started. See Pricing and credits for how credits are consumed. For a custom volume arrangement, contact customer.success@kardinal.ai.