Service · Agent Kits

Six named AI agents, deployed in 4 to 6 weeks, for mission-driven work.

Agent Kits are productized Copilot Studio agent deployments for nonprofits, foundations, and rural hospitals. Pick the agent that fits the workflow you most want to take off your team's plate. We deliver it governed, in production, in 4 to 6 weeks.

Talk to us about scope See the six kits →

4 to 6 weeks · $15,000 to $25,000 fixed price · One agent in production · No commitment beyond delivery

Why Agent Kits

The barrier is not the technology. It is doing it well.

Mission-driven teams are at capacity. The mission is not slowing down. Microsoft calls the answer "digital labour": AI agents that handle specific tasks at human direction, 24/7, without forgetting processes or taking institutional knowledge with them when they leave.

Copilot Studio lets you build an agent in a weekend. The barrier is scoping the right first agent, designing it so it produces real outcomes rather than a chatbot demo, wiring it into your data and identity controls so it can be trusted with mission-critical work, and training the team to use it.

Most organizations try to build an agent themselves, get one working in pilot, and stall before it produces measurable value. The Kit format compresses the work into a four-to-six-week engagement with a fixed scope, a fixed price, and a single named agent in production at the end.

Pick the Kit that maps to the workflow you most want to reclaim.

The six Agent Kits

Six patterns that recur across the mission sector.

Each Kit ends with one named agent live, integrated with your Microsoft 365 and Copilot Studio environment, governed under Entra Agent ID, and operating against documented KPIs.

A two-row, three-column catalog of the six Agent Kits (Donor Intelligence, Intake and Routing, Knowledge Keeper, Research Assistant, Constituent Support, and Board Reporter), each showing target user, capability, Microsoft stack, and fixed price.

Donor Intelligence Agent Kit · $20,000 · 5 weeks

Donor intelligence agent

For development teams at nonprofits and foundations. Prioritizes prospects from your CRM, drafts personalized outreach grounded in donor history and wealth signals, tracks every touchpoint, surfaces stalled relationships before they go cold.

The team gets back: fundraiser time spent in conversations with people, not in spreadsheets.

Governance: sensitivity labels on donor PII, access scoped to development staff, audit trail on every draft.

Microsoft stack: Copilot Studio, Microsoft 365 Copilot, Dataverse, Microsoft Graph connectors.

Intake and Routing Agent Kit · $20,000 · 5 weeks

Intake and routing agent

For nonprofits, legal aid organizations, foundations with grant operations, and hospital case-management teams. Processes incoming forms, extracts data, classifies cases by type and urgency, routes to the right staff, drafts the intake summary, files supporting documents.

The team gets back: a three-hour intake queue compressed to roughly twenty minutes. Eligibility errors caught early.

Governance: PII detection, sensitivity labels propagated to drafts, human-in-the-loop confirmation before send.

Microsoft stack: Copilot Studio, Power Automate, Forms, SharePoint, Dataverse.

Knowledge Keeper Agent Kit · $15,000 · 4 weeks

Knowledge keeper agent

For any mission-driven organization that has lost institutional knowledge through staff turnover. Maintains organizational knowledge, answers staff questions about policies, procedures, donor histories, grant requirements, and clinical protocols. Surfaces stale or contradictory content to a human reviewer.

The team gets back: institutional memory that does not walk out the door.

Governance: retrieval grounded only in approved SharePoint sources, sensitivity labels respected on responses.

Microsoft stack: Copilot Studio, SharePoint agents, semantic index, Microsoft Graph.

Research Assistant Agent Kit · $20,000 · 5 weeks

Research assistant agent

For program officers, grant writers, policy teams, and hospital clinical leads. Consultant-quality research on grants, programs, policy, and public data, pulled from vetted sources you specify, cited inline, drafted into the format you actually need: executive memo, grant background section, board paper, or policy brief.

The team gets back: consultant-grade research output without the consultant fees.

Governance: source citations on every output, external sources allowlisted, audit trail for grant compliance.

Microsoft stack: Copilot Studio, Microsoft 365 Copilot with grounding, Bing Custom Search connectors, Azure AI Foundry.

Constituent Support Agent Kit · $25,000 · 6 weeks

Constituent support agent

For nonprofits with constituent-facing programs, foundations with grantee-facing operations, and rural hospitals with patient-facing intake. 24/7 self-service for common questions: eligibility, application status, hours, documentation requirements, intake links. Logs every interaction. Escalates anything outside its scope to a human with a written summary.

The team gets back: fewer repetitive tickets, faster answers for the community, the after-hours coverage a lean staff cannot provide on its own.

Governance: explicit escalation triggers, human-in-the-loop on anything ambiguous, bilingual where required.

Microsoft stack: Copilot Studio, Microsoft Teams, Power Automate, Microsoft 365.

Priced higher because of multi-channel deployment and external-facing governance.

Board Reporter Agent Kit · $20,000 · 5 weeks

Board reporter agent

For executive directors, foundation program officers, hospital CFOs, and anyone who builds quarterly reports. Synthesizes program data, financials, and impact metrics into board-ready materials on demand. Drafts the quarterly narrative. Updates dashboards. Catches numbers that look off and flags them for review.

The team gets back: quarterly reports drafted in the time it used to take to assemble the data. Board questions answered between meetings.

Governance: source-of-truth data only (Microsoft Fabric, SharePoint, Dynamics), attribution on every number, human edit before send.

Microsoft stack: Copilot Studio, Microsoft Fabric, Power BI, Microsoft 365 Copilot.

How it works

Six phases. The same engineering discipline across every Kit.

The agent is different for each engagement. The methodology is not.

A horizontal six-step methodology shared across every Agent Kit: Discovery (week 1), Agent Design (week 2), Build (weeks 2–4), Pilot, Adoption Training, and Handoff, with AgentOps discipline built in throughout.

  1. 01

    Discovery

    Week 1. Workflow mapping, data-source inventory, success criteria documented with the workflow owner.

  2. 02

    Agent design

    Week 2. Conversation design, grounding strategy, integration design, governance design (Entra Agent ID, data-access scope, audit logging). Design review with the workflow owner.

  3. 03

    Build

    Weeks 2 to 4. Agent built in Copilot Studio. Integrations wired. Test data loaded. Internal QA.

  4. 04

    Pilot

    Weeks 4 to 5. Live with a pilot group of 5 to 15 users. Daily check-ins. Iteration based on real use.

  5. 05

    Adoption training

    Week 5. Role-based training for the team that owns the workflow. Documentation handed off.

  6. 06

    Handoff

    Week 6. Operating runbook delivered. Adoption metrics baseline captured. 60-minute leadership presentation. Every Kit ships with AgentOps discipline built in: version control, performance monitoring, drift detection, rollback procedures.

Pricing at a glance.

Fixed price. One agent in production at the end of each engagement. No 12-month commitment.

Agent Kits: duration and fixed price per engagement.
Agent Kit Duration Fixed price
Donor Intelligence 5 weeks $20,000
Intake and Routing 5 weeks $20,000
Knowledge Keeper 4 weeks $15,000
Research Assistant 5 weeks $20,000
Constituent Support 6 weeks $25,000
Board Reporter 5 weeks $20,000

Microsoft 365 Copilot and Copilot Studio licensing is client-procured (passthrough). Licensing is not included in the Kit price.

What happens after the Kit.

Every Kit ends with one agent in production and a clear choice for what comes next.

Continue alone for a quarter

The agent is live, the team is trained, the runbook is in your hands. Operate it yourselves and re-engage Centered Networks when you want the second agent or a review.

Stack another Kit

Some clients deploy two or three Kits over a year before graduating to CompleteCare Intelligence. The Kits are designed to compose: Donor Intelligence and Board Reporter are a common pair; Intake and Routing and Knowledge Keeper are another.

Graduate to CompleteCare Intelligence

The ongoing managed AI program. Intelligence includes a monthly 20-engineering-hour pod for new agent builds, AgentOps discipline at scale, FinOps for AI, and quarterly board-ready ROI reporting. Most clients who deploy more than one agent graduate to Intelligence. See CompleteCare Intelligence →

Agent Kits run on CompleteCare Automate, the Power Platform tier of CompleteCare that handles workflow automation, Power Automate connectors, and agent infrastructure. See CompleteCare Automate →

Who Agent Kits are for.

Good fit

  • Organizations on CompleteCare Foundations (or equivalent baseline) with Microsoft 365 Copilot already licensed or in active procurement
  • Teams with a clear, workflow-specific problem the agent will own
  • Mission-driven organizations with 50 to 1,000 users
  • Leadership that wants a single named agent live and useful, not a multi-quarter platform initiative

Not a fit

  • Organizations that have not yet decided whether Copilot is the right investment. Start with a Discovery Sprint.
  • Organizations whose data governance is not yet in place. The agent inherits whatever governance your tenant has; if Foundations is not yet running, that is the prerequisite work.
  • Teams looking for a chatbot demo. Agent Kits build production agents with measurable KPIs. If you want a quick prototype, Copilot Studio does that on its own.

Rural or critical-access hospital?

Kits can be delivered HIPAA-aligned for rural hospitals. Start at Healthcare AI Readiness to confirm fit before scoping a Kit.

When custom is the right answer

Some workflows do not fit a Kit.

For those, we build custom on Microsoft Azure OpenAI Service and Azure AI Foundry, with the same governance posture.

  • Claims Denial Navigator For rural and critical-access hospitals: AI that parses 835 denial files, identifies appeal opportunities, and drafts appeals. Eligible for the Microsoft Rural Hospital Resiliency Program. See Healthcare AI Readiness →
  • Document processing at volume High-volume intelligent extraction, parsing, and routing of forms and correspondence, when the Kit pattern is not enough and Azure Document Intelligence is the right tool.
  • Custom Azure AI applications When the workflow requires a custom application on Microsoft Azure AI Foundry, with data residency, content safety, and audit posture configured to your sector.

Questions

Frequently asked questions.

How do we choose which Kit to start with?

On the scoping call we walk through the six and identify which workflow has the strongest ROI signal for your team right now. If two are tied, we usually recommend starting with the one your senior workflow owner is most excited about. Adoption matters more than theoretical ROI in the first agent.

Why these six?

They map to the six highest-ROI agent patterns we see across the mission-driven sector. Microsoft publishes general agent scenarios (HR, IT, customer service, sales, marketing, e-commerce). We have repackaged those patterns into nonprofit, foundation, and healthcare workflows that match the work mission-driven teams actually do.

Can you build a custom agent that is not one of the six?

Yes. That is a custom engagement, not an Agent Kit. We scope it as a one-off statement of work or fold it into a CompleteCare Intelligence engagement. Agent Kits are the productized version because most teams' first agent fits one of the six patterns.

Do we need to be on CompleteCare Foundations first?

You need the baseline: Conditional Access, sensitivity labels, MDM, Entra ID governance. If you are already there, you can run an Agent Kit standalone. If not, Foundations is the prerequisite; we will stage the Kit 60 to 90 days out.

What does the agent run on after handoff?

Your Microsoft 365 and Copilot Studio environment. The agent lives in your tenant. The runbook documents the maintenance pattern. AgentOps is on you unless you graduate to CompleteCare Intelligence, where we run AgentOps for you.

How is this different from buying an off-the-shelf agent from another vendor?

Two differences. First, the agent is grounded in your data, your governance, your identity controls, not a tenant-shared SaaS that sees data from other customers. Second, you own the agent at the end. There is no per-user-per-month fee paid to a third party. The agent runs on your Microsoft licensing.

What is Entra Agent ID and why does it matter?

Entra Agent ID is Microsoft's identity governance layer for AI agents. Every agent gets a managed identity, scoped permissions, and audit logging. It is what makes the agent treatable as a digital teammate with all the access governance that implies, rather than as an anonymous bot. Every Kit deploys with Entra Agent ID configured.

What if the agent does not produce the KPIs we hoped for?

Each Kit includes a pilot phase specifically to catch this. If the pilot shows the agent is not going to hit the KPIs, we iterate. If iteration does not close the gap, the Handoff Document is candid about what is possible and what is not. We will never ship an agent into production we would not run ourselves.

The No-Lock-In Promise mark: month-to-month from day one, no 12-month contract, no termination fee, with a 30-day exit chip.

Choose an Agent Kit.

One named agent. Four to six weeks. Fixed price. In production at the end.

Tell us which Kit interests you and the workflow you have in mind. We confirm fit on a 30-minute call and get the engagement on the calendar.

Not sure which fits? Start with a Discovery Sprint → The roadmap tells you which agent has the strongest ROI signal for your team.

This field is required
Valid email required
This field is required

Thanks, we’ve got it.

A senior member of our team will reach out within one business day to set up the scoping call.

Built on Microsoft

Data and AI designated partner.

Agent Kits are built on Microsoft Copilot Studio, governed through Microsoft Entra Agent ID, and grounded in your Microsoft 365 and Dataverse environment. No proprietary platform. No third-party data processor. The agent lives in your tenant from Day 1.

Microsoft Solutions Partner for Data and AI (Azure) designation badge. Microsoft Solutions Partner for Modern Work designation badge.
  • Microsoft Solutions Partner for Data and AI (Azure)
  • Microsoft Solutions Partner for Modern Work
  • Entra Agent ID governance on every Kit
  • Agents grounded in your tenant data
  • AgentOps discipline included
  • HIPAA-aligned delivery available for healthcare