Kaizo provides two ways to automate quality assurance using SamurAI (Kaizo’s AI): Auto QA and Autopilot.

Both features evaluate customer conversations, but they differ in how ratings are generated and the level of human involvement required.

🔍 Quick comparison 

🤝 Auto QA

Auto QA is a hybrid QA workflow where AI assists QA experts by pre-filling specific scorecard criteria.

When a QA expert opens a conversation for evaluation, SamurAI analyzes the interaction and pre-selects the automated criteria based on your instructions. The reviewer can confirm or override the AI suggestion before completing the rating.

When to use Auto QA

Use Auto QA when:

• ⚙️ You want to automate simple or objective criteria (for example, greeting, grammar, or sign-off)

• 👀 Your team wants human oversight on every evaluation

• 🧪 You are testing AI-assisted QA before enabling full automation

⚠️ Important: When using Auto QA, QA Autopilot should remain disabled.

🤖 Autopilot

Autopilot enables fully automated QA evaluations.

SamurAI automatically evaluates resolved conversations in the background using the configured scorecard. The results appear in QA House, where they can be reviewed if needed.

When to use Autopilot

Use Autopilot when:

• 📊 Your AI instructions are well tested

• 🧠 All scorecard criteria are automated

• 🚀 You want to increase QA coverage across all conversations

By default, Autopilot evaluates 5% of resolved conversations, but this percentage can be adjusted in the Cases Funnel settings depending on your plan.

📌 Key takeaway

• Auto QA assists QA experts by pre-filling selected scorecard criteria during manual reviews.

• Autopilot automatically evaluates conversations without requiring manual QA.

Many teams start with Auto QA to refine their AI instructions before enabling Autopilot for full QA automation.