Skip to content
NewEvery agent tool call now routes through the governed MCP gateway — policy check, audit log, then dispatch.Read the changelogarrow_forward
XOOZLYAI Orchestrator
routeHow it workssmart_toyAgent fleetshieldGovernance & trustsellPricingmenu_bookDocsarticleBloghistoryChangelog
Log inStart free
grid_viewProductsrouteHow it workssmart_toyAgent fleetshieldGovernance & trustsellPricingmenu_bookDocsarticleBloghistoryChangelogHelp centreLog in
Start free
hubGovernance & Orchestration core hubverified_userSOC 2 Type II · ISO 27001 · ISO 42001

The control plane for enterprise-grade AI agent fleets.

Orchestrate multi-agent delivery with continuous guardrails, hard cost caps and an auditable trail behind every tool call. Take a product specification to production code without losing governance along the way.

Start freearrow_forwardshieldExplore the trust centre
Specialized agents
61
Core domains
10
SDLC stations + hub
8 + 1
Tool calls audited
100%
app.xoozly — orchestration loop8 stations synchronised · core hub routing
hubXOOZLYCore Intelligence
  1. searchAStrategy & Research
  2. paletteBDesign
  3. architectureCArchitecture
  4. codeDDevelopment
  5. bug_reportEQA
  6. rocket_launchFDevOps & Release
  7. trending_upGGrowth & Marketing
  8. forumHPost-Launch & Feedback
  1. searchAStrategy & Research
  2. paletteBDesign
  3. architectureCArchitecture
  4. codeDDevelopment
  5. bug_reportEQA
  6. rocket_launchFDevOps & Release
  7. trending_upGGrowth & Marketing
  8. forumHPost-Launch & Feedback
  9. gavelHubGovernance & Orchestration
shield_lockDeterministic guardrailsEvery tool call is policy-checked and audited before it leaves the platform.
Stations A–H run plan → execute → evaluate → feedback, then loop back to A. Category I governs every hand-off between them.

Enterprise orchestration

Why enterprise fleets need a governed control plane

Ad-hoc scripts and unversioned agent loops create token spend leaks, security vulnerabilities and legal exposure. XOOZLY wraps autonomous agents in deterministic, enforceable controls.

  • shield_lock

    Deterministic guardrails

    Every agent tool call passes one choke point — policy check, audit log, then connector dispatch. No domain can bypass it, so a misconfigured permission on one agent cannot cascade through the fleet.

    • check_circleMCP tool gateway with per-agent, per-project scope grants
    • check_circleAppend-only audit record for every call, allowed or denied
  • savings

    Hard cost caps and FinOps

    A project's cost cap is a hard limit checked before a run starts, not reconciled after it ends. Spend is tracked per span, and a breach pauses the run instead of invoicing you for it.

    • check_circleMulti-provider routing by task class, not a single vendor
    • check_circleSpan-level spend tracking with auto-pause on breach
  • how_to_reg

    Human-in-the-loop control

    The Rule Optimizer proposes; a person approves. Approval writes a new versioned rule file rather than mutating the old one, so every agent's behaviour stays reviewable and rollback-ready.

    • check_circleApproval gates on the actions you choose, not all or nothing
    • check_circleVersioned rule files with actor attribution on every change

Closed-loop architecture

The orchestration loop in action

Agents operate as specialized stations around the core hub. Every cycle leaves its own requirements trail: planned, executed, evaluated, fed back and optimized.

  1. 01Station A

    searchStrategy & Research

    Agents that research markets, competitors, and requirements before work begins.

    Requirements · research briefs→ B
  2. 02Station B

    paletteDesign

    Agents that produce UI/UX designs, wireframes, and visual specifications.

    Tokens · mockups→ C
  3. 03Station C

    architectureArchitecture

    Agents that define system architecture, data models, and technical design decisions.

    ADRs · schema→ D
  4. 04Station D

    codeDevelopment

    Agents that write, test, and ship application code.

    Commits · pull requests→ E
  5. 05Station E

    bug_reportQA

    Agents that test builds, find regressions, and verify quality before release.

    Test runs · regressions→ F
  6. 06Station F

    rocket_launchDevOps & Release

    Agents that build, deploy, and release infrastructure and application changes.

    Releases · rollbacks→ G
  7. 07Station G

    trending_upGrowth & Marketing

    Agents that run campaigns, analyze funnels, and drive user acquisition.

    Campaigns · funnels→ H
  8. 08Station H

    forumPost-Launch & Feedback

    Agents that monitor production, gather user feedback, and drive iteration after launch.

    Feedback · rule proposals→ A
hubContinuous execution cycle

Anchored by the Governance & Orchestration core hub

Category I routes between every station and holds the approval gates; Category J, Platform Intelligence, runs across all of them. Station H loops straight back to A, so feedback re-enters planning instead of ending in a report nobody reads.

Start your first looparrow_forward

Complete coverage

61 specialized agents. 10 core domains.

Every agent ships with a versioned rule file, a declared tool list and a risk tier. A declaration is never a grant: an agent can only call a tool a gateway policy has explicitly allowed in that project.

  • searchCategory A

    Strategy & Research

    Agents that research markets, competitors, and requirements before work begins.

    4 agentschevron_right

  • paletteCategory B

    Design

    Agents that produce UI/UX designs, wireframes, and visual specifications.

    3 agentschevron_right

  • architectureCategory C

    Architecture

    Agents that define system architecture, data models, and technical design decisions.

    4 agentschevron_right

  • codeCategory D

    Development

    Agents that write, test, and ship application code.

    9 agentschevron_right

  • bug_reportCategory E

    QA

    Agents that test builds, find regressions, and verify quality before release.

    9 agentschevron_right

  • rocket_launchCategory F

    DevOps & Release

    Agents that build, deploy, and release infrastructure and application changes.

    9 agentschevron_right

  • trending_upCategory G

    Growth & Marketing

    Agents that run campaigns, analyze funnels, and drive user acquisition.

    6 agentschevron_right

  • forumCategory H

    Post-Launch & Feedback

    Agents that monitor production, gather user feedback, and drive iteration after launch.

    4 agentschevron_right

  • gavelCategory I

    Governance & Orchestration

    Agents that enforce rules, review compliance, and coordinate the rest of the fleet.

    7 agentschevron_right

  • psychologyCategory J

    Platform Intelligence

    Agents that analyze platform-wide data and improve how the fleet itself operates.

    6 agentschevron_right

Start freearrow_forwardBrowse the full marketplacearrow_forward

Enterprise trust and compliance

Rigorous compliance built into the runtime

Data is encrypted at rest with AES-256 and in transit with TLS 1.3. Standard tiers carry a 99.9% uptime SLA; Enterprise deployments carry 99.95%.

  • verified_user

    SOC 2 Type II

    Independently audited controls over security, availability and confidentiality.

    Certified · annual audit

  • lock

    ISO 27001

    An information security management system covering the platform and its operations.

    Enterprise grade

  • smart_toy

    ISO 42001 (AI Systems)

    The AI management standard, applied to how the fleet itself is governed and improved.

    Active governance posture

  • policy

    GDPR · CCPA · DPDP

    Data subject requests and legal holds are first-class operations, not a support ticket.

    DSAR intake · legal holds

Audit logs are append-only, and no rule file version or invoice is ever mutated in place. Visit the Trust & Security Centrearrow_forward

Transparent model

Predictable per-seat licensing with transparent usage

Scale your agent capacity without unpredictable overage penalties. Pick a project ceiling, not a token budget — model spend stays bounded by your own cost caps on every tier, including the free one.

  • Developer Free Tier

    Self-serve

    For individual engineers evaluating orchestration loops.

    Per-seat access

    Bring your own Anthropic, OpenAI, Gemini or OpenRouter key — you pay the model directly, with no markup from us.

    • check_circleUp to 3 concurrent projects
    • check_circleThe full 61-agent catalogue, all 10 domains
    • check_circleThe 8-station orchestration loop end to end
    • check_circleGoverned MCP tool calls — policy check, audit log, dispatch
    • check_circlePer-project cost caps with auto-pause on breach
    • check_circleAppend-only audit log and execution traces
    Start free
  • Most popular

    Team Control Plane

    Team scale

    For teams running multi-agent loops in production.

    Per-seat + usage-based

    A pooled monthly allowance, bounded by the same hard cost caps as every other tier.

    • check_circleUp to 25 concurrent projects
    • Everything in Developer, plus
    • check_circleHuman-in-the-loop approval queue with Slack hand-off
    • check_circleRule Optimizer proposals and versioned rule files
    • check_circleSpan-level cost reporting and spend forecasting
    • check_circleGitHub, Slack and Stitch connectors
    • check_circleWorkflow automation, webhooks and scheduled reports
    • check_circleTOTP multi-factor auth and Google SSO
    Get started with Team

  • Enterprise Dedicated

    Custom

    For regulated organisations with strict data-residency needs.

    Custom contract terms

    Invoicing, MSA and SLA guarantees agreed per organisation and recorded against your org.

    • check_circleUnlimited concurrent projects
    • Everything in Team, plus
    • check_circleCustom roles and permission groups
    • check_circleRow-level tenant isolation across every project
    • check_circleDSAR fulfilment, legal holds and RoPA tooling
    • check_circleOrg impersonation with a full audit trail
    • check_circle99.95% uptime SLA and priority incident response
Explore the full plan matrix and feature comparisonarrow_forward

Every tier enforces the same governance: no agent reaches a tool except through the gateway, and no run starts above its cost cap.

XOOZLYAI Orchestrator

Ready to orchestrate your AI fleet with total confidence?

Start a project, set a cost cap, and watch the loop run with every tool call policy-checked and logged before it leaves the platform.

Start freearrow_forwardSee what shipped
  • check_circleNo credit card required
  • check_circleYour keys, your model spend
  • check_circleEvery tool call audited

Platform

  • All productsarrow_forward
  • Orchestration looparrow_forward
  • Agent fleetarrow_forward
  • Governed control planearrow_forward
  • Pricingarrow_forward

Trust & legal

  • Trust centrearrow_forward
  • Security & compliancearrow_forward
  • AI use policyarrow_forward
  • Report a vulnerabilityarrow_forward
  • Data subject requestsarrow_forward

Resources

  • Documentation & guidesarrow_forward
  • Engineering & research blogarrow_forward
  • Changelogarrow_forward
  • Uptime & reliabilityarrow_forward

Browse by domain

  • Strategy & Researcharrow_forward
  • Developmentarrow_forward
  • QAarrow_forward
  • Governance & Orchestrationarrow_forward
  • All 10 domainsarrow_forward
XOOZLY
© 2026 XOOZLY · 61 agents · 10 domains · 8 SDLC stationsAI use policyarrow_forwardLog inarrow_forward99.9% uptime SLAarrow_forward