Autonomous Sourcing

Systems that run sourcing events with limited human intervention: configuring and launching requests for quotation, collecting and scoring supplier bids, and optimising award allocation across lots and constraints. They operate during the sourcing event itself and are commonly applied to repeatable and standardised categories.

Lifecycle position: Sourcing event
Unit of work: Sourcing event
Typical buyer: Sourcing managers and operational procurement teams handling repeatable events.
Representative tools: Keelvar, Arkestro, Pactum

How to recognise this class

Ingests requirement specifications, bidder lists, lot structures and award constraints for a defined event.
Outputs launched requests for quotation, scored supplier bids and optimised award allocations across lots.
Runs during the sourcing event, replacing manual configuration and evaluation steps with automated ones.

Commonly confused with

Scenario-Based Procurement Analytics — Autonomous sourcing configures and runs the sourcing event itself, while scenario-based procurement analytics operates earlier, modelling the commercial outcome of a planned contract before any event is launched.

Where it sits in the cycle

Preceded by Scenario-Based Procurement Analytics (Pre-award).
Followed by Transactional & Intake AI (Transaction execution).

Part of the full classification of AI systems in procurement.