HOME/TECHNIQUE/Agentic Orchestration/Routing / intent classification

TECHNIQUE

Routing / intent classification

Agentic Orchestration

3APPLICATIONS
5OBSERVED OPERATORS
01

State of Practice

CROSS-VALIDATED — 9 OPERATORS

Routing / intent classification is deployed as a control-plane decision: operators use it to choose agents, retrieval paths, backends, widgets, workflow branches, reviewers, or human experts before committing downstream work.

Observed Practices

Use an explicit routing or intent-classification step to choose the next downstream path rather than treating the model output as the final product.

9 of 9 operators in the deployed/pilot roster are cited with routing or intent-classification evidence.
AmazonAudibleDoorDashDropboxLinkedInRexeraRipplingUberWix

Route early to scope the context, retrieval strategy, domain skill, or review profile before deeper processing.

4 of 9 operators are cited using routing to narrow context, retrieval, domain, or sub-agent scope.
DoorDashDropboxLinkedInRippling

Use routing to select among specialist agents or assistants instead of sending every request through one generalist path.

4 of 9 operators are cited splitting work across specialists with a lead scout, classifier, supervisor, or pluggable assistant framework.
DoorDashDropboxRipplingUber

Gate or suppress outputs using confidence, category, or value filters after classification.

3 of 9 operators are cited using confidence-based filtering or category suppression around routed/classified outputs.
AmazonAudibleUber

Classify customer search intent to decide which product/search surface to show, including visual widgets.

2 of 9 operators are cited using customer intent detection in visual autocomplete.
AmazonAudible

Use reward-based routing to assign operational work to a human expert.

1 of 9 operators is cited routing tickets by expected reward over possible experts.
Wix

Where Operators Converge

Every cited operator uses routing / intent classification as an intermediate orchestration decision that changes what system component runs next: agent, assistant, backend, graph node, review profile, widget, branch, or expert.

Where Operators Diverge

Operators differ on what the router selects.

APPROACH 01

Select a specialized agent, sub-agent, or assistant.

DropboxRipplingUber

APPROACH 02

Select a retrieval strategy, backend, graph node type, or memory layer.

LinkedIn

APPROACH 03

Select a customer-facing surface or human expert.

AmazonAudibleWix

APPROACH 04

Select a domain review profile or workflow branch.

DoorDashRexera

Operators differ on the decision mechanism used for routing.

APPROACH 01

Supervisor, LLM, tool-calling, or function-calling based routing.

DropboxLinkedInRippling

APPROACH 02

Learned, probabilistic, ranking, or reward-scoring models.

AmazonAudibleWix

APPROACH 03

Workflow branches or domain profiles that constrain the path.

DoorDashRexera

APPROACH 04

Post-generation category classifier that tags and suppresses classes of output.

Uber

Watch Items

Wrong routing in complex workflows can produce false positives or false negatives; Rexera explicitly reports agents taking the wrong path and veering off course, while Amazon and Audible add confidence filters for ambiguous general-vs-specific queries.

Noisy or low-value classified outputs hurt user trust: DoorDash says noisy generic reviewers get ignored, Uber says simple standalone prompts create many false-positive comments, and Rexera ties off-course agents to false positives and false negatives.

Routing logic needs continuous validation as systems and data change: Dropbox warns that changes to intent classification and adjacent stages can ripple into hallucinations, and Wix calls out regular updates plus heavy validation for integrity and data drift.

02

Implementation Menu

CURATED DEFAULTS
NameKindMaturity
Constrained JSON-schema classification on a fast modelpatterncommodity
Embedding-similarity routingpatternestablished
03

Observed in Production

3 APPS