# Bedrock AgentCore vs Amazon Quick Suite — enterprise decision matrix

Use this matrix during architecture review. **Amazon Quick Suite** is AWS’s forward path for net-new enterprise assistant buyers (July 2026); existing **Amazon Q Business** customers may continue on Q Business or migrate indexes to Quick Suite.

> **Reproduce this** — Copy scenario rows into your RFC. Pair with [`tco-worksheet.csv`](tco-worksheet.csv) for monthly cost assumptions.

## Quick routing (5 questions)

| # | Question | If yes → lean Quick Suite | If yes → lean AgentCore |
|---|----------|---------------------------|-------------------------|
| 1 | Are users employees or analysts with SSO? | ✓ | |
| 2 | Do you need permission-aware search over SharePoint/Confluence/S3 without custom code? | ✓ | |
| 3 | Is the UX AWS-managed chat (web, Teams, Slack)? | ✓ | |
| 4 | Is the agent customer-facing or embedded in your product API? | | ✓ |
| 5 | Do you need custom orchestration (LangGraph, multi-agent, MCP tools you own)? | | ✓ |

## Scenario matrix

| Scenario | Quick Suite | AgentCore | Hybrid | Why |
|----------|:-----------:|:---------:|:------:|-----|
| Internal developer assistant (docs + runbooks) | ✓ | | | Connectors + ACL inheritance; Q Developer for IDE separately |
| Employee productivity (HR/IT policies) | ✓ | | | Managed RAG + admin console |
| Customer support (external users) | | ✓ | | Custom auth, branding, consumption economics |
| IT help desk (internal) | ✓ | | | ServiceNow/Jira connectors; avoid building RAG |
| Knowledge management | ✓ | | | Quick Index + 40+ connectors |
| Enterprise search | ✓ | | | Permission-aware unified index |
| Business automation (no-code) | ✓ | | | Quick Flows / Quick Automate |
| Custom AI agents (product feature) | | ✓ | | Runtime + Gateway + Memory |
| Autonomous multi-step workflows | | ✓ | | Harness or LangGraph on Runtime |
| Document processing pipeline | | ✓ | | Bedrock Data Automation + AgentCore tools |
| Research assistant (internal) | ✓ | | | Quick Research + trusted sources |
| Data analysis (governed BI) | ✓ | | | Quick Sight on curated datasets |
| Software engineering (IDE) | — | — | ✓ | **Q Developer** (third lane); not AgentCore |
| Platform engineering (infra automation) | | ✓ | | AgentCore + Lambda/API tools |
| Finance (policy Q&A) | ✓ | | | ACL-sensitive documents |
| HR (benefits, policies) | ✓ | | | Document-level permissions critical |
| Sales (CRM + decks) | ✓ | | | Salesforce connector + Quick Sight |
| Marketing (content drafts) | ✓ | | | Quick Suite; not production ad agents |
| Healthcare (PHI internal Q&A) | ✓ | | | BAA-eligible Q/Quick; audit via CloudTrail |
| Manufacturing (SOPs, safety) | ✓ | | | Shop-floor knowledge via mobile/web |
| Government (FedRAMP path) | ✓ | | | Region residency; GovCloud scope per AWS |
| Embedded public FAQ bot | ✓ | ✓ | | Quick consumption units vs AgentCore API — model TCO |
| High-volume transactional agent (millions/mo) | | ✓ | | Consumption vs seat economics invert |

**Legend:** ✓ = primary recommendation; Hybrid = both (e.g., Quick Suite internal + AgentCore customer-facing).

## When NOT to choose Quick Suite

| Situation | Choose instead |
|-----------|----------------|
| White-label product chat with your UX | AgentCore + Bedrock Converse API |
| Sub-second latency SLA on first token | AgentCore with warm pools; not index-sync chat |
| Complex tool chains you must version in Git | AgentCore Runtime (code-based agent) |
| Anonymous high-traffic public embed at scale | AgentCore — Quick consumption bundles spike (see TCO worksheet scenario B) |

## When NOT to choose AgentCore

| Situation | Choose instead |
|-----------|----------------|
| IT wants AI helpdesk in 30 days, no engineers | Quick Suite |
| Only need “ask our wiki” for 500 employees | Quick Suite Lite/Pro + Enterprise Index |
| Team has no agent ops maturity | Quick Suite first; AgentCore after one production agent |

## Related

- [Bedrock AgentCore vs Amazon Q decision guide](/blog/aws-bedrock-agentcore-vs-amazon-q-enterprise-decision-guide-2026/)
- [AgentCore production architecture](/blog/amazon-bedrock-agentcore-production/)
- [Amazon Q for Business setup](/blog/how-to-set-up-amazon-q-for-business-sharepoint-s3/)