Managed AI services & Copilot consulting: Your team is already using AI. We’ll make it safe, and worth it.

Owning IT is in our nature.

You get Copilot or Claude set up properly, your people trained, your data kept where it belongs, and regular reports on who’s using it and what it costs.

Make your AI safe and useful → See what it costs

Technical specs & details ↓ Or start with a free IT health check →

On every ManageIT plan: a live expert in 5 minutes, or a $100 credit · See your price before you call

Behind: RITIS Productivity, the reporting included with AmplifyAI — sample environment shown.

In short

AmplifyAI is DivergeIT’s managed AI service for Microsoft Copilot, Anthropic Claude, Google Gemini and Azure AI Foundry: it governs which AI tools your people use, supports them, and shows you adoption and cost. AI Success Desk plans start from $1,000 a month (Plus from $1,900), and Forward Deployed Engineering is scoped from $14,800. Your prompts and files never pass through DivergeIT systems. Full specs · Pricing

The problem

It’s not just you. AI showed up before anyone wrote the rules.

Your team is already using it. The question is whether anyone is running it.

Shadow AI

People are using AI tools you never approved, sometimes with client data.

Licenses nobody uses

You bought Copilot seats. A few people tried it once.

No idea if it’s paying off

Everyone says you need AI. Nobody can show you who’s using it or what it costs.

Nobody owns it

IT thinks it’s a business decision. The business thinks it’s an IT project.

Heard on first calls

“Half my team is pasting client info into ChatGPT, and I can’t stop them.”
“I bought Copilot licenses. Nobody uses them.”
“Everyone says I need AI. Nobody can tell me what for.”

The real issue

The problem isn’t the AI. It’s that nobody is running it.

AI tools are easy to buy and easy to ignore. Without someone setting rules, training people and measuring use, they turn into either a risk or a line item.

AmplifyAI is one team behind everything AI touches: governance, training, support, and reports on who’s using it and what it costs.

Find out where you stand →

Three coworkers gathered around one laptop at a sunny shared table, trying out a new tool together

So where do you start?

Red to green, in three steps.

First — Govern

We set the rules

Approved tools, clear usage rules and data protection, so client information stays where it belongs.

Days 1–30 — Onboard

We get your team using it

Onboarding and training in 30 days, built around the work your people actually do.

Ongoing — Measure

We show you how it’s used

Reports show who’s using it and what it costs, plus shadow-AI detection, reviewed with you quarterly on Core or monthly on Plus.

Make your AI safe and useful →

Before & after

What changes for you

Before: Staff use whatever AI tool they find.

With DivergeIT: Your team gets approved AI tools and clear rules, and unapproved tools get spotted.

Before: Licenses sit unused.

With DivergeIT: Your team learns it in 30 days, and you see who’s using it.

Before: No idea if it’s worth the money.

With DivergeIT: Reports show who’s using it and what it costs, reviewed with you quarterly on Core or monthly on Plus.

What's included

Inside AmplifyAI

AmplifyAI is one AI Success Desk to manage it all. Not just a ticket queue — a dedicated team that governs, supports, and grows your AI investment across Microsoft Copilot, Anthropic Claude, Google Gemini, and Azure AI Foundry: adoption reporting, shadow-AI detection, cost visibility, and a skill library your people actually use. And when you are ready to build, Forward Deployed Engineers embed with your team.

Every engagement starts with a 30-day onboarding: audit the environment, set the acceptable-use policy, roll out pilot-first, light up reporting, then your AI Success Desk goes live.

AI Success Desk

White-glove support for your designated tools, including Copilot and Claude, with best practices, a support response target of one business day, and business reviews quarterly on Core or monthly on Plus.

Forward Deployed Engineers

An AI engineer embedded with your team — maps your workflows, builds the agents they need, then hands them back managed.

Adoption & usage reporting

Active users, session depth, and adoption by team — plus token and usage cost visibility by user and team.

Shadow-AI detection

See unsanctioned AI use across your organization, with tool blocking and enforcement on AI Success Desk Plus when detection is not enough.

Skill library & custom skills

Pre-built skill and prompt library with marketplace plugin and connector support — plus, on Plus, custom skill generation refined against your real workflows.

30-day onboarding

Audit, govern, roll out, measure, activate: platform configuration, M365 data-risk review, enforced AI acceptable-use policy, and pilot-first rollout.

Explained

Managed AI services and Copilot consulting, in plain terms

What are managed AI services?

Managed AI services means an outside team runs your AI the way a managed IT provider runs the rest of your IT: it governs which tools your people use, supports them, measures adoption and cost, and maintains the agents you deploy. You get the productivity without unmanaged risk or runaway spend.

AmplifyAI does this for Microsoft Copilot, Anthropic Claude, Google Gemini and Azure AI Foundry. Every engagement starts with a permissions and shadow-AI sweep, then your AI Success Desk keeps adoption, cost and policy on track.

Productivity. Adoption that sticks: usage reporting by team, re-engagement for people who stopped, and a skill library built around your workflows.
Cost. License and token spend visible by user and team, with budget alerts and cost controls on Plus.
Risk. Shadow AI found and reported, and on Plus, unapproved tools blocked at the network edge.

What does Microsoft Copilot consulting include?

Good Copilot consulting starts before the licenses: a review of who can see what in Microsoft 365, because Copilot surfaces anything a user already has access to. Then an acceptable-use policy enforced in your tenant, a pilot-first rollout, and reporting on who actually uses it.

With AmplifyAI, that work continues after launch through your AI Success Desk, and when you’re ready, a Forward Deployed Engineer builds agents around your real workflows and hands them back managed.

See the AmplifyAI spec sheet →

RITIS · Productivity

Adoption you can see, not adoption you assume

The hardest question about an AI rollout is whether anyone is using it. RITIS counts real activity per licensed person across Email, Teams, Files and Copilot, which is how AmplifyAI reports on whether the seats you bought turned into habits.

The Productivity module. Adoption per service, counted per person. Sample environment shown.

What the Productivity screen scores

Email 99%

93 of 94 active · 30 days
Teams 78%

73 of 94 active · 30 days
Files 80%

75 of 94 active · 30 days
Copilot 17%

15 active · 16 ever used

See the Productivity module → Try it on your data

Also reported here: Support. All five domains roll into one RITIS Score.

Technical specs & details

AmplifyAI, line by line

Exactly what each AI Success Desk tier covers across productivity, risk and cost, which platforms you can run, and how your data is handled.

Platforms · Microsoft Copilot · Anthropic Claude · Google Gemini · Azure AI Foundry

Tiers at a glance

AI Success Desk Core

“See it”: a dashboard on your AI and a help desk your people can call.

Minimum: Covers your first 10 users

Adoption & usage reporting by user and team
Shadow-AI detection & reporting
Token and usage cost visibility
Designated Copilot / Claude support & best practices
Skill, prompt & training libraries
Support response target: 1 business day · quarterly reviews

AI Success Desk Plus

“Control it”: everything in Core, and now you act on it.

Minimum: Covers your first 10 users

Everything in Core, plus —
Tool blocking for unapproved AI
Role-based permissions management
Cost controls & budget alerts
Custom skill generation with your users
DevOps / GitHub setup for developer teams
One templated agent build per quarter · monthly reviews

Feature matrix

Capability	Core	Plus
Productivity
Adoption & usage reporting (active users, session length & frequency)	Included	Included
Inactive-user re-engagement	Included	Included
Designated-tool support & best practices (Claude / Copilot)	Included	Included
Pre-built skill library and curated prompt library	Included	Included
Marketplace plugin & connector support	Included	Included
General AI training library KnowBe4	Included	Included
General tool use: chat, cowork, code	Included	Included
Risk
Shadow-AI detection & reporting (unsanctioned app usage)	Included	Included
Tool blocking & enforcement Web Defense	Not included	Included
AI policy review & governance cadence	Not included	Included
Role-based permissions management	Not included	Included
Cost
Token & usage cost visibility (spend by user and team)	Included	Included
Cost controls & budget alerts	Not included	Included
Service
Support response target	1 business day	1 business day
Business review cadence	Quarterly	Monthly
Custom skill generation & refinement with your users	Not included	Included
DevOps / GitHub administration & setup	Not included	Included

Add-on services

Managed agents

Per agent, monthly

Every deployed agent monitored and maintained like any other system
Builds from Forward Deployed Engineering land on this meter
An agent is anything registered with an Entra Agent ID

Forward Deployed Engineer

Hourly, or a scoped engagement

An AI engineer embedded with your team
Maps your workflows and builds the agents for them
Hands every build back managed, on your agent meter
With an AI Success Desk plan: two weeks in, results that show it paying for itself within a year, or your Forward Deployed Engineering fees back

Requirements & what’s not included

Engagement

Sold on a 12-month initial term as an add-on to a DivergeIT managed-services agreement.
Role-based permissions management needs an admin license for each AI product.
Every engagement starts with a permissions and shadow-AI sweep of your tenant.
Custom connectors, agents or automations DivergeIT builds or hosts are separate engagements, each with its own statement of work.

Your data & privacy

Is DivergeIT a sub-processor of our AI data?

No. A sub-processor is a company your AI vendor engages to process your data on its behalf. DivergeIT has no such role with Microsoft, OpenAI, Anthropic or Google. DivergeIT works for you, under your agreement.

Does our AI data pass through DivergeIT systems?

No. Prompts, files and responses travel directly between your users and the AI platform, stored in your workspace under the vendor’s terms. AmplifyAI does not proxy, copy or retain that content.

Why does DivergeIT need admin access?

To set up SSO, configure tenant and security settings, publish your approved-tools policy, manage licenses, connect approved data sources, and support your users. Admin access changes settings; it is never a feed of your content. It uses named accounts you approve and can remove at any time.

Will DivergeIT staff see our prompts or outputs?

Not as part of the service. Engineers see AI content only when your users share it to get help, such as on a screen share or in a ticket, and anything shared is covered by your agreement’s confidentiality terms. Audit logs and DLP events stay in your platform.

How does blocking unapproved AI tools work?

AmplifyAI uses Web Defense to keep users on your approved tools by blocking unsanctioned AI sites at the network edge. It is part of your DivergeIT security services and governed by your DivergeIT agreement.

DivergeIT is SOC 2 Type 2 certified by AICPA. These answers describe how the service is delivered; your agreement governs.

Want this as a document for your team? Book a call and get the full service definitions →

Pricing

Every rate we charge is published in one place

AmplifyAI tiers, what each one includes, and the assumptions behind the numbers — on the pricing page, not scattered across the site. On ManageIT, call or chat and you reach a live expert within 5 minutes, or you get a $100 credit.

See pricing →

The guarantees

Promises with consequences, not fine print.

Eligibility & limits → Read the agreements →

Response

On any ManageIT plan, wait more than 5 minutes for a live expert and you get a $100 credit on your next invoice.

All ManageIT plans

Ransomware recovery

Hit by ransomware anyway? Your covered data and servers are restored to their pre-attack state at no DivergeIT charge.

ManageIT Plus & Pro, with DivergeIT immutable backup

Switching

If switching disrupts your operations, your first 30 days of service are credited in full.

New ManageIT agreements

Satisfaction

If your concern still isn’t resolved, you get a $50 gift of your choice.

All ManageIT plans

Annual assessment

Your assessment and prioritized recommendations are free, every year. Any fixes are quoted separately.

Any SecureIT plan (included in every ManageIT plan)

AI payback

Not on track to pay for itself after the first two weeks? You get your Forward Deployed Engineering fees back.

Forward Deployed Engineering with an AI Success Desk plan

What changed for them

“The team produced an amazing AI tool that is going to maximize our team's efficiency.”

Gary R. · CEO, Creative Design Group More client stories →

FAQ

AI Solutions questions, answered

What is the AI Success Desk?

One team that governs, supports, and grows your AI investment across Microsoft Copilot, Anthropic Claude, Google Gemini, and Azure AI Foundry. Core covers adoption and usage reporting, designated-tool support, a skill and prompt library, shadow-AI detection, and cost visibility, with a support response target of one business day. Plus adds tool blocking, policy and permissions management, cost controls, and custom skills. It is a success desk, not just a help desk, and it works alongside the Forward Deployed Engineers who build for you.

What do Forward Deployed Engineers actually do?

An AI engineer embeds with your team, maps your workflows, and builds the agents they need — then hands them back managed. Every build becomes a managed agent, monitored and maintained at a flat per-agent monthly rate. Work is hourly or a scoped engagement.

What do the first 30 days look like?

Five steps: audit your platform configuration and data risk, set an AI acceptable-use policy enforced in tenant settings, roll out pilot-first with the AmplifyAI skill library, light up usage and spend reporting, and go live with your AI Success Desk.

Copilot, Claude, Gemini, or Azure AI Foundry — which one?

Your AI Success Desk supports Microsoft Copilot, Anthropic Claude, Google Gemini, and Azure AI Foundry. We recommend by fit against your workflows, data estate, and budget, and the desk, skill library, and governance model work the same whichever you choose.

What is a managed agent?

An AI agent we built or onboarded that runs in your environment under AI Success Desk management — monitored, maintained, and metered at a flat per-agent monthly rate rather than by usage surprises.

How we run it

AmplifyAI, built for your industry

Professional Services Entertainment & Sports Financial Services All industries

Related services