Import application APIs from rayorch:

API	Purpose
Pipeline	declare a static dataflow in forward() and execute one finite input with pipeline.run(...)
RayModule	declare a UDF, constructor args, outputs, and resources
function	adapt a callable into a stateless RayModule
F.expand, F.filter, F.broadcast, F.reduce	declare cardinality and lineage
run	equivalent functional form: rayorch.run(pipeline, ...)
Executor	reuse persistent actor pools across runs
RunResult, OutputIssue, ItemOutcome	inspect outputs and run metrics
RecordFailure, GroupFailure	return explicit business failures from UDFs
RecoveryPolicy	configure bounded per-Call recovery

Import experiment APIs from rayorch.benchmark:

API	Purpose
built-in *Bench classes	typed Benchmark configurations
available, load, register	lazy discovery and registration
BenchmarkReport	serializable config, metrics, profile, artifacts, outputs
LocalSource	map a source checkout to Ray runtime environment fields
BenchmarkRun	status, logs, stop, and wait for a submitted Ray Job

Compatibility boundary

Modules named _program, _runtime, _execution, _model, and _protocol are implementation details. Avoid importing their types from application code. Public APIs remain small so internal scheduling and storage representations can evolve.

For signatures and current behavior, package source and docstrings are authoritative while the project is in alpha.

Contributors: Sunnyhaze