Waves
Waves = execution units
A wave represents a self-contained and self-sufficient group of work that delivers tested and accepted functionality.
Waves are how the flow principle manifests in practice --> the visible execution units where dependency-driven work actually happens. They can be seen as the marriage of sprints (delivery focus) and epics (business focus), optimized for AI-assisted development.
The wave concept replaces time-boxed sprints, since artificial time boundaries don't make sense when code can be implemented in hours.
Waves characteristics:
A wave contains all dependencies needed for execution --> self sufficiency
All tasks within an epic must be in the same wave
Length of the wave is less important. 2 hours or 2 days, that's the same.
Waves complete as cohesive units with delivered capabilities - atomicity
Waves Types
Hydro categorizes work into four natural wave types that follow the logical progression of software development. Each wave type has distinct characteristics in duration, dependencies, and AI-human collaboration patterns.
Foundation Waves - Establish the architectural groundwork that enables all subsequent development. Could include any platform or data related work.
core architecture
shared components
development environment
Feature Waves - Build user-facing capabilities and business logic on top of the foundation
user-facing capabilities
business rules
data processing
Integration Waves - Connect individual features into cohesive system experiences
system cohesion
end-to-end workflows
cross-feature integration
Polish Waves - Optimize and enhance the quality of completed functionality. Includes bug fixing.
performance tuning
documentation
user experience enhancement
Foundation
1-2 days (emergent)
None or minimal external requirements
70% AI-ready tasks (interfaces, schemas, configurations)
Enables all subsequent feature development
Feature
2-3 days (emergent)
Foundation waves plus previous feature waves
40% AI-ready, 40% AI-assisted, 20% human-only
Delivers measurable business value to users
Integration
3-5 days (emergent)
Multiple feature waves
20% AI-ready, 40% AI-assisted, 40% human-only
Production-ready system capabilities
Polish
1-2 days (emergent)
Core functionality complete
50% AI-ready (refactoring, documentation), 50% human-only (UX, optimization)
Enterprise-grade quality and maintainability
Wave completion checkmarks
A wave is considered complete when all validation gates pass:
✔︎ All tasks within wave closed and validated ✔︎ Human operator validation complete ✔︎ All code committed to version control ✔︎ Integration tests passing at wave boundary ✔︎ Next wave dependencies satisfied and unblocked ✔︎ Performance benchmarks met ✔︎ Security scans completed (if applicable) ✔︎ Documentation updated and accurate ✔︎ Business acceptance criteria validated
Post-Completion activities:
Automatic dependency analysis identifies newly unblocked tasks
Next wave activation when team ready and dependencies satisfied
Cascade analysis triggers to determine priority of ready tasks within active wave
Metrics collection for methodology improvement and business reporting
This creates comprehensive quality gates that prevent technical debt cascade while enabling natural flow progression to subsequent waves.
Last updated