Process Optimization

Improve how work gets done.

What This Skill Does

Analyzes business processes and workflows to identify inefficiencies, eliminate waste, and design optimized processes. Uses techniques like value stream mapping, bottleneck analysis, and continuous improvement principles.

When to Use

Activate this skill when:

• User wants to improve a process or workflow
• Phrases like "optimize this", "be more efficient", "workflow improvement"
• Analyzing how work currently gets done
• Identifying and removing process bottlenecks
• User says "this process is broken" or "how can we streamline"

How It Works

The optimization process ensures systematic improvement:

1. Map current state - Document the existing process
2. Identify issues - Find bottlenecks, waste, pain points
3. Analyze root causes - Understand why problems exist
4. Design future state - Create optimized process
5. Plan implementation - Define changes and rollout
6. Measure impact - Track improvements

Input Parameters

Parameter	Type	Required	Description
process_name	string	Yes	Name of process to optimize
scope	string	No	Process boundaries and scope
stakeholders	list	No	People involved in the process
metrics	object	No	Current performance metrics

Process Analysis Techniques

Value Stream Mapping

Identify value-added vs. non-value-added activities:

Activity Type	Definition	Action
Value-Added	Customer is willing to pay for	Keep and optimize
Non-Value Added	Necessary but no direct value	Minimize
Waste	No value, not necessary	Eliminate

The 8 Wastes (DOWNTIME)

Waste	Description	Examples
Defects	Mistakes requiring rework	Bugs, errors, rework
Overproduction	Creating more than needed	Unused features, excess reports
Waiting	Delays between steps	Approval queues, dependencies
Non-utilized talent	Not using people's skills	Manual work that could be automated
Transportation	Moving things unnecessarily	File transfers, handoffs
Inventory	Hoarding work in progress	Large backlogs, stale branches
Motion	Unnecessary movement	Switching between tools
Extra processing	Doing more than required	Redundant approvals

Output Structure

The skill generates optimization outputs:

1. Current state analysis - Documentation of existing process
2. Problem identification - List of issues with root causes
3. Future state design - Optimized process definition
4. Implementation plan - Steps to achieve optimization

Analysis Format

# Process Optimization: [Process Name]

## Current State

### Process Map

[Step 1] → [Step 2] → [Step 3] → [Step 4] → [Step 5] ↓ ↓ ↓ ↓ ↓ (2 min) (1 day) (2 min) (3 days) (5 min)

### Metrics
| Metric | Current Value | Target |
|:-------|:--------------|:-------|
| Cycle time | 5 days | 1 day |
| Touch points | 5 | 3 |
| Rework rate | 30% | <5% |
| Satisfaction | 4/10 | 8/10 |

### Stakeholders
| Role | Involvement | Pain Points |
|:-----|:-----------|:------------|
| Role 1 | Steps 1, 4 | [Their complaints] |
| Role 2 | Steps 2, 3, 5 | [Their complaints] |

## Problems Identified

| Problem | Impact | Root Cause | Type |
|:--------|:-------|:-----------|:-----|
| Approval bottleneck | +3 days | Single approver | Waiting |
| Rework from errors | +1 day | Unclear requirements | Defects |
| Manual data entry | +30 min | No integration | Non-value added |

## Future State

### Optimized Process Map

[Step 1] → [Step 2] → [Step 3] ↓ ↓ ↓ (2 min) (1 hour) (5 min)

### Changes
| Change | Impact | Effort |
|:-------|:-------|:-------|
| Parallel approvals | -2 days | Medium |
| Automated validation | -1 day | Low |
| Remove redundant step | -30 min | Low |

### Expected Metrics
| Metric | Current | Target | Improvement |
|:-------|:--------|:-------|:------------|
| Cycle time | 5 days | 1 day | 80% |
| Touch points | 5 | 3 | 40% |
| Rework rate | 30% | <5% | 83% |

## Implementation Plan

### Phase 1: Quick Wins (Week 1)
- [ ] Remove redundant approval step
- [ ] Add automated validation
- [ ] Create process documentation

### Phase 2: System Changes (Week 2-3)
- [ ] Build integration between systems
- [ ] Implement parallel approval workflow
- [ ] Update UI for new process

### Phase 3: Training & Rollout (Week 4)
- [ ] Train stakeholders on new process
- [ ] Monitor and adjust
- [ ] Measure results

## Risks & Mitigation
| Risk | Probability | Impact | Mitigation |
|:-----|:------------|:-------|:-----------|
| Resistance to change | Medium | Medium | Involve stakeholders early |
| Technical issues | Low | High | Pilot before rollout |
| Process breaks | Low | High | Maintain rollback option |

Quality Standards

Before delivering, the optimization should:

• Quantify current and future performance
• Identify specific root causes, not symptoms
• Propose realistic, implementable solutions
• Consider stakeholder perspectives
• Include measurement plan
• Balance automation with human judgment

Collaboration Modes

Mode	How It Works
autopilot	Analyzes process, generates optimization plan
copilot	Identifies issues, validates solutions with you
manual	Provides analysis frameworks, you complete

Context Integration

Reads:

• context/feedback-synthesis.json - User pain points and complaints
• context/impact-analysis.json - Process performance data

Writes:

• context/process-optimization.md - Analysis and recommendations

Read By:

• prd-gen - Incorporates process improvements into requirements
• iteration-planning - Includes optimization work in sprints

Example Usage

User: "Our code review process takes too long"
→ Analyzes current process, identifies bottlenecks, proposes improvements

User: "Optimize the customer onboarding workflow"
→ Maps current state, designs streamlined future state

User: "Why does deployment take so many steps?"
→ Identifies waste, creates implementation plan for CI/CD improvement

Best Practices

1. Map before changing - Understand current state first
2. Go to gemba - Observe the process where it happens
3. Ask why 5 times - Find root causes, not symptoms
4. Start with quick wins - Build momentum with easy fixes
5. Measure everything - You can't improve what you don't measure