Six-Stage AI Lifecycle (Hong Kong)
Tags: Frameworks
TL;DR
- Six stages align AI delivery with SDLC: Strategy, Planning, Ecosystem, Development, Deployment, Operation & Monitoring.
- Emphasises data work and iterative feedback between development, deployment, and operations.
Why it matters for HK marketers: It provides a project spine to integrate AI into martech stacks with the right checkpoints.
The six stages
- 1. Project Strategy: Organisation strategy, internal policies/practices, industry standards and regulations.
- 2. Project Planning: Portfolio management, project oversight, delivery approach.
- 3. Project Ecosystem: Technology roadmap for AI/data, procuring AI services (sourcing), change management.
- 4. Project Development: Business & data understanding, solution design, data extraction and pre-processing, model building.
- 5. System Deployment: Model integration & impact, transition & execution (with failure mitigation), ongoing monitoring, evaluation & check-in (traceability, repeatability, reproducibility).
- 6. System Operation and Monitoring: Data & model performance monitoring, operational support, continuous review/compliance.
Alignment with SDLC
- Maps to standard SDLC stages from project request to maintenance.
- Recognises continuous data sourcing and iterative model training beyond one-off builds.
6 lifecycle stages align AI with traditional SDLC.
So what for marketers
Plan AI initiatives against these stages and lock in data, testing, and monitoring plans early to maintain accuracy and trust post-launch.
← Back to Knowledge Base