Cloud, security & AI, delivered like it runs in production
Citadel designs, secures, and operates production systems for enterprises, with fixed-price consulting and governed managed AI agents. Hire the architects who run this work every day.
What Citadel delivers for enterprises, and on what terms
Citadel Cloud Management delivers architect-led cloud, security, and AI work for enterprises across five areas: cloud architecture and review on AWS, Azure, and GCP; security and GRC aligned to SOC 2, ISO 27001, HIPAA, and FedRAMP controls; zero-trust identity and access governance; AI enablement from proof of concept through to governed production; and scoped 6R migration assessments with cost and risk called out before a commitment. Commercial terms are fixed-price under a written statement of work, typically inside a master services agreement for ongoing work, priced after a 30-minute discovery call rather than metered by the hour. Consulting engagements typically run from $10,000 to more than $500,000 depending on scope. The published entry point is the Agent Assurance Assessment, three weeks at $15,000. Ongoing work graduates into monthly retainers, starting at $8,000 a month for Managed AI Operations. Deliverables and the configuration built for the client environment belong to the client.
Pricing
Fixed-price engagements, typically $10K to $500K+
Depending on scope. Priced after a discovery call: no open-ended time-and-materials meter, no surprise invoices. You know what a first phase costs before you commit.
Start here
Agent Assurance Assessment: 3 weeks, $15,000, price published
The fixed-scope front door to everything above. An inventory and risk map of the AI agents already running in your environment, a governance evidence pack mapped to NIST AI RMF and ISO/IEC 42001, one working governed agent deployed in your environment, and an executive readout with a prioritized rollout plan. A larger multi-environment tier is published at $40,000. No “contact us for a quote”.
A project is the start, not the end. Most engagements graduate into an ongoing retainer, so your governed AI keeps running, improving, and staying audit-ready. Monthly, scoped after discovery.
Managed AI Operations
From $8,000 / mo
Run, monitor, and improve the agents you deployed.
Operation, monitoring, and tuning of your governed agents
Autonomy graduated as trust builds (T1 → T2 → T3)
Monthly action-log review + SLA-backed where the engagement calls for it
Most chosen
Fractional Chief AI Officer
$8,000 to $25,000 / mo
Senior AI leadership without a full-time executive hire.
Owns AI strategy, architecture direction, and vendor decisions
Board- and exec-ready roadmaps and risk framing
Steers your delivery teams and governance model
AI Governance & Compliance
Retainer
Stay audit-ready as your AI footprint grows.
Continuous control checks + evidence assembly (SOC 2 / ISO 27001 / HIPAA / FedRAMP-aligned)
Per-agent identity, autonomy tiers, and audit trails maintained
From first call to operating in production: four steps, fixed-price, no black boxes.
1
Discovery (30 min)
A working call to map your stack, constraints, and the highest-value first step. No pitch deck.
2
Scoped fixed-price proposal
A written SOW with deliverables, milestones, and a fixed price, no open-ended time-and-materials meter.
3
Delivery
Senior architects execute against milestones, with governance applied to every AI-agent action from day one.
4
Operate / handoff
Monitored operation or a clean handoff with runbooks, SLA-backed where the engagement calls for it.
AI Agents · SIAS
506 agents, one governance model
Software In A Service: production-ready agents across 8 domains: engineering, security, data, ops, and more. Each agent is its own principal with scoped credentials and full action logging, run at the autonomy tier you set. Designed, secured, hosted, and monitored by Citadel.
Explore SIAS The path every agent action takes. The autonomy tier decides who approves it; the audit log means nobody has to take the answer on trust.
Representative engagements
Composites of how Citadel scopes and governs work, not named clients. Named outcomes are published here as engagements clear for reference.
Representative engagement
SOC triage, isolation human-gated
Context:
Mid-size SaaS security team drowning in phishing and alert volume.
Problem:
Analysts spent most of their day on repetitive triage, leaving real incidents waiting.
What Citadel did:
Deployed Tier-1 triage and evidence-collection agents; every isolate/disable action left at Tier 3.
Governance:
Full action logging, reversible writes, four-eyes on any containment step.
Outcome:
Routine-triage analyst time materially reduced while containment stayed human-owned.
Representative engagement
Fixed-price cloud migration, 6R assessment
Context:
Platform team facing an aging estate and an open-ended migration quote from another vendor.
Problem:
No clear view of cost, risk, or sequencing before committing budget.
What Citadel did:
Delivered a scoped 6R assessment with cost and risk called out per workload, against a fixed price.
Governance:
Fixed-price SOW with milestone sign-off; no time-and-materials meter.
Outcome:
A defensible migration plan the team could take to their own leadership for approval.
Why Citadel
Fixed-price scope
Engagements are scoped and priced after discovery, no open-ended time-and-materials meter.
Governed by design
Every AI agent is a first-class principal: its own identity, scoped credentials, full action logging.
Multi-cloud, senior-led
Delivered by working architects across AWS, Azure, and GCP, not a bench of juniors.
Citadel vs the alternatives
The honest trade-offs against a Big-4 consultancy, staff augmentation, and building it yourself.
Comparison of delivery models
Citadel
Big-4 consultancy
Staff augmentation
DIY / generic agent vendor
Who does the work
Working architects
Junior bench, senior sales
Contractors you manage
Your team, stretched
Pricing
Fixed-price, up front
Time & materials
Hourly per head
Opportunity cost
AI-agent governance
First-class identity + T1 to T3 + full audit
Varies by team
Not included
Build it yourself
Speed to value
Scoped first step in weeks
Long ramp
Ramp per hire
Whenever capacity frees up
Lock-in
Runbooks + clean handoff
Retainer pull
Knowledge walks out
None
Procurement & delivery questions
The legal, data, and contract questions that decide a $250K engagement, answered up front.
Where do the agents run and who holds our data?
Agents run in or adjacent to your tenancy, grounded in your corpora, which stay under your control and are tenancy-isolated per engagement. See the security & governance page for the full posture.
Is the deployment single-tenant?
Yes. Corpora and agent identities are isolated per engagement, and privileged corpora are isolated per legal domain so one workstream cannot read another.
Who owns the agents and IP we build together?
Deliverables and the configuration built for your environment are yours. The specific IP and license terms are set in the engagement agreement so your legal team reviews them before signing.
What contract type do you work under: fixed-price, SOW, MSA?
Consulting engagements are fixed-price under a written SOW, typically inside an MSA for ongoing work. Pricing is set after a discovery call, not metered by the hour.
What happens to our corpora when we stop?
Grounding data and issued credentials are decommissioned and access is revoked at exit, with retention and deletion terms defined in the agreement up front.
What if an agent gets it wrong on a write?
Writes happen only at supervised (T2) or human-gated (T3) tiers within scoped credentials. Every action is logged, attributable, and designed to be reversible, so a mistake is traceable and can be rolled back.
How fast do we see first value?
Most engagements start with a scoped, lower-risk first step, often a Tier-1 read-and-report agent or a fixed-price assessment, so there is a concrete deliverable early rather than a long ramp.