Stage 1: discover
This module helps business leaders, project managers and team leaders identify and evaluate automation opportunities to create a scalable, value-based automation pipeline.
Involve relevant multidisciplinary stakeholders when identifying and prioritising automation tasks. Your organisation’s process owners, users, customers, technology and data experts, policy and legal teams, and automation experts will have valuable contributions.
Automation opportunities
The aim of Stage 1: discover is to identify, prioritise, and capture automation opportunities in the master backlog or ‘pipeline’. First, understand existing processes and workflows to identify areas where automation may bring value.
Use the questionnaire below to identify processes in your organisation’s automation pipeline:
1. Create the backlog by using the task automation suitability checklist
| Task checklist | Description |
| Is it very manual? | The task involves mundane and manual work. |
| Is it rule-based and clear process (stable and consistent)? | The process is structured with clear, standardised, and predictive decision-making instructions. |
| Has it low rate of exceptions? | The task has low number of variations. |
| Is it highly repetitive? | The task is highly frequent and has high volume of transactions. |
| Is it prone to human error? | The task has high rate of human error. |
| Is it mature and measurable? | The process is stable, predictable, well-documented and known operational costs. |
| Are its inputs standard and digitally readable? | The process inputs triggered by digitally readable inputs such as Excel, Word, Email, PDF. |
| Are systems and processing methods remaining unchanged long-term? | The process does not require change in its fundamental technical architecture or processing methods. |
| Does automation save timing costs? | The task when automated saves the time of at least 2 to 3 full-time employees (FTE). |
2. Finalise the backlog by considering the business and use case context
| Consider | Detail |
| Data preparation | Consider AI model and data suitability. Address any gaps, inconsistencies, bias or limitation impacting the prototype, final solution, and the chosen AI model. |
| Data sensitivity | Does the task or process involve handling sensitive or confidential data? If so, additional security, compliance, data ownership and procurement measures may be mandated. |
| Task complexity | Determine the complexity of the task. Highly complex tasks may not be suitable for automation without significant resources. |
| AI suitability | Consider the appropriate AI component, model or algorithm based on use case, data availability and desired outcomes. |
| Cost of automation | Consider the upfront and ongoing costs of implementing and maintaining automation, including software, hardware, personnel and, AI models. |
| Human intervention | Decide if the task requires human intervention or decision-making. Consider legal advice and Human-in-The-Loop (HITL) for solution oversight and failsafe mechanisms. |
| Legal and regulatory compliance | Ensure that the task or process complies with all relevant laws and regulations. Automation must not violate any legal requirements. |
| Impact on community | Ensure the task or process is human-centred. The implemented solution should be inclusive, respects human rights, diversity, and autonomy of individuals. |
| Impact on employees | Assess the potential impact on employees. If automation replaces human workers, consider retraining or redeployment plans. |
| Integration with existing systems | Verify if the task or process can be integrated seamlessly with existing systems and technologies. |
| Return on investment | Calculate the expected ROI for automation. Automation benefits such as time and cost savings, should outweigh the implementation costs. |
| Scalability or volume | Consider whether the task or process needs to scale up or down based on business needs. Scalability may influence the choice of automation tools. |
| Testing and quality assurance | Determine how you will test and ensure the quality of automated processes. Robust testing and quality control are essential for success. |
Automation pipeline prioritisation
Prioritise opportunities according to their strategic alignment to your organisation’s goals and your business context.
Automation opportunities will be prioritised according to
Automation opportunities will be prioritised according to:
- Their strategic alignment to your organisation’s goals
- The suitability of tasks for automation to organisational context including:
- Business impact: impact on cost savings, cost avoidance, ROI or efficiency gains
- Feasibility: technical complexity, resource availability and faster implementation time
- Process criticality: core operational processes, and high volume and frequency
- Compliance and risk: improving regulatory compliance or mitigating significant risk
- Stakeholder desirability: impacted parties buy-in and executive sponsorship
- Data availability and quality: quality data readiness and ease of integration.
Steps to build an automation pipeline from your backlog based on business value
Steps to build an automation pipeline from your backlog based on business value:
- Define the outcome for each use-case:
- each use-case should have a measurable metric aligned with business goals.
- consider the increased amount of oversight and due diligence if the use case requires use of AI creating potential harm.
- Define the business value for each use-cases:
- define measures for efficacy, efficiency and agility based on customer improvements, cost, or risk outcomes.
- Categorise the automation use-cases:
- task automation: deliver quick benefit through automating repetitive manual tasks.
- process automation: deliver high business value through workflows and business processes orchestration and automation.
- augmentation: empower employees with intelligent capabilities.
Next lifecycle stage
Stage 2: build
Design and build automation solutions by producing a comprehensive implementation plan.