← Back to Mantleway
Mantleway Readiness Assessment

A working agent.
On your real project files.
In two weeks.

A fixed-price, fixed-scope engagement that audits your file server, validates your Microsoft 365 setup, and stands up a working Copilot agent grounded in your real project archive. No migration. No restructuring. No multi-month commitment. Everything we produce is yours to keep, whether or not you continue with us.

Investment $5,500 Fixed fee. No change orders.
Duration 2 weeks Active work. Up to two on-site visits.
What You Keep Everything The agent, the documentation, the runbook — yours whether you continue or not.
If We Find Blockers We Pause Build credit carries forward. No additional cost.
Book the Assessment → See the 2-week plan
01 The Plan

What actually happens in those two weeks.

  1. Day 01–03

    File Server Audit

    Two engineers, one to two days on your floor. We catalogue what's actually in your project archive, surface the permission complexity hiding inside your AD groups, and identify the structural issues that will block a Copilot rollout if left alone.

    • File-share inventory across the project archive — formats, age distribution, depth, and orphaned content
    • NTFS permission analysis — group sprawl, broken inheritance, and the directories where ACLs no longer match reality
    • Identity model audit — Active Directory hygiene, security group consistency, account churn against current staff
    • Stakeholder interviews with three to five engineers, focused on what they actually try to find and how they currently fail
    Day 03 Output File Server Readiness Report — data inventory, permission complexity rating, identity model assessment.
  2. Day 04–06

    Microsoft 365 Validation

    We verify your tenant is actually configured to do what Microsoft's marketing implies it can do. Identity bridging, license fitness, capacity envelope — checked against the real shape of your file server, not a theoretical reference architecture.

    • Entra Connect (formerly Azure AD Connect) hybrid identity verification — sync health, attribute mapping, conflict resolution
    • Copilot license inventory and seat-fit assessment against the engineers who would actually use the system
    • Microsoft Graph Connector capacity envelope — index size estimate against your archive, throttling exposure, refresh cadence
    • Tenant configuration review — DLP policies, conditional access, retention rules that may interfere with grounding
    Day 06 Output Microsoft 365 Configuration Assessment — gaps named, prerequisite work scoped, ready-to-go-or-not verdict.
  3. Day 07–11

    Workflow Analysis & Agent Build

    We pick one workflow that matters — RFI response, submittal review, proposal drafting, or basis-of-design recall — and stand up a working Copilot agent grounded in your real project files. Not a demo on synthetic data. Your archive, your engineers, your problems.

    • Workflow scoring across the four candidate domains, with the firm's leadership choosing the target
    • Microsoft Graph Connector deployed against the relevant file shares, ACL mirroring verified before any indexing
    • Agent configuration with citation requirements baked in — no answers without source-document references
    • Test run against ten real historical questions from the chosen workflow, with answer-quality scoring
    Day 11 Output One working Copilot agent deployed to your tenant, queryable by your engineers against your real project files.
  4. Day 12–14

    Readout & Implementation Proposal

    We deliver the documentation package, walk your leadership through findings in person, and hand over a fixed-fee implementation proposal — with no obligation to proceed. The agent stays. The runbook stays. The decision is yours.

    • Executive Readout delivered in person (your office or ours) to the principal and any stakeholders you invite
    • Fixed-fee implementation proposal scoped against the four-stage Mantleway Method (Audit → Bridge → Build → Wire)
    • Knowledge transfer to your IT lead — agent administration, refresh procedures, monitoring, and known limitations
    • Honest conversation about what we recommend — including the workflows we don't think Copilot is ready to handle yet
    Final Output A working agent, three written documents, and a fixed-fee proposal you can take to your principal — or to a different vendor.
02 The Deliverable

What the Readiness Report looks like.

Every Readiness Assessment ends with this document. Below is a redacted preview from a recent engagement — the real thing is longer and specific to your environment.

MANTLEWAY READINESS ASSESSMENT · FILE SERVER READINESS REPORT

[Client Firm] · MEP Engineering · [State]

Engagement RA-[NNN]

Executive Summary

Over 14 days, Mantleway audited [████] of project-archive content across [████] active and archived projects, validated the firm's Microsoft 365 hybrid identity configuration, and deployed a working Copilot agent grounded in the firm's RFI history. The agent passed [████]/10 historical-question tests with citations, with [████] identified prerequisite items to be resolved before broader rollout.

What We Audited

Archive size[████] TB across [████] shares
Identity modelHybrid · Entra Connect
NTFS group count[████] groups, [████] orphaned
Workflow chosenRFI Intelligence

Findings

  • Proven Microsoft Graph Connector successfully indexed the [████] active-projects share with verified ACL mirroring against existing AD groups. Engineers see exactly what their NTFS permissions allow.
  • Caution [████] permission inconsistencies identified in the archived-projects share — broken inheritance, group references no longer in AD, and at least three directories where the ACL no longer matches stated firm policy. Remediation recommended before broader index expansion.
  • Do Not The [████] share contains client-confidential drawings governed by NDAs that prohibit AI processing. We recommend explicit exclusion from the connector and a separate governance conversation before any agent has access.

Recommended Next Phases

  1. Phase 1Permission remediation on the archived-projects share (est. 3 weeks, [████])
  2. Phase 2Build & deploy three additional agents — Submittal Review, Proposal Drafting, BOD Recall (est. 6 weeks, [████])
  3. Phase 3Wire agents into Outlook, Teams, and Pipedrive via Power Automate (est. 4 weeks, [████])
Prepared by Mantleway · [Month] 2026 Delivered in person · [City], [State]

The real document is typically 20–30 pages across three reports (File Server Readiness, Microsoft 365 Configuration, Top Workflows Analysis), with appendices covering the agent test set, identity-model details, and connector capacity projections. You receive the complete package in print and PDF on Day 14, regardless of whether you engage us further.

03 Common Questions

What procurement and security teams tend to ask.

Q1 We don't have Microsoft 365 Copilot licenses yet. Should we still book this?
Yes — and arguably it's the right time. Without licenses we focus on Microsoft 365 readiness and a license-fit recommendation, which is more honest than buying first and discovering later that your tenant won't support what you need. If you do have at least a handful of Copilot licenses to test against, the Day 7–11 agent build runs on those. We'll tell you in the intake call which path applies.
Q2 Where will our project files actually go? Are you copying them to the cloud?
Your files stay on your file server. The Microsoft Graph Connector indexes content — it builds a searchable representation of your archive that lives inside your Microsoft 365 tenant — but it does not lift the files to a different location. Agents query the index. We do not copy, snapshot, or retain your data, and neither does Microsoft outside your tenant boundary. If your IT team wants to verify this, we'll walk them through the connector configuration before any indexing starts.
Q3 What does your security review look like from our side?
We meet your perimeter — we do not ask for exceptions. The connector deploys inside the security posture your file server already carries, and we mirror your existing NTFS permissions before any agent has access. We complete your vendor intake process during Day 01–03, provide insurance documentation on request, and are happy to have your IT lead review every step of the configuration before it touches production. Background checks on our engineers available on request.
Q4 What if you find blockers we didn't know about — bad permissions, identity issues, license gaps?
We deliver the assessment portion (the Readiness Reports) and pause the agent build until prerequisite work is resolved. The fee does not change. The build credit carries forward — when blockers are fixed, you get the agent build at the original price, with no expiration date attached to that credit. The findings document tells you exactly what needs to happen, who typically does it, and roughly how much it costs. You can have us do the remediation, or your existing IT vendor, or your own team.
Q5 What happens on Day 15 if we want to continue?
You receive a fixed-fee implementation proposal during the Day 12–14 readout — covering the four-stage Mantleway Method (Audit, Bridge, Build, Wire) at a price calibrated to what we found in your environment. There is no "convert to annual" pressure. The proposal is yours to take to your principal, your board, or a competing vendor. If you decide not to proceed, the agent and the documentation are still yours. Many firms run the agent for months on the assessment alone before committing to broader work.
Q6 Can we extend the two weeks?
The 14-day window is deliberate — it forces focus, it protects your budget, and it matches the attention span of most principals trying to make a decision about AI spend. We do not extend the assessment. If the work surfaces complexity beyond the scope, we complete what we can in 14 days and recommend a follow-on engagement at transparent pricing. Scope does not silently grow.
Q7 Will Copilot actually work for us, or is this just hype?
The honest answer is: it depends on your archive, your permissions, and the specific workflow you're trying to support. Copilot grounded in a clean, well-permissioned project archive against a citation-required workflow (RFI response, submittal review, BOD recall) is genuinely useful. Copilot pointed at a messy share with broken permissions and asked to do open-ended creative work is not. The Readiness Assessment exists to tell you which category you're in before you spend more.
Q8 How many engagements are you running at once?
Deliberately few. Mantleway is a small practice by design, and the principal is on every engagement. If calendar availability is tight, we'll tell you when you inquire — we do not overbook.
Book the Assessment

Two weeks is a short time to learn a lot.
We like it that way.

Send a short note. We'll reply within one business day with a proposed start window and a simple intake form. No sales call unless you ask for one.

Start Here Book the Assessment Opens a pre-drafted email with the five fields we need to schedule. hello@mantleway.com → Not Ready Yet Ask a question first Honest questions get honest answers. Usually within a business day. Ask by email →
If We Find Blockers, We Pause — No Additional Cost

If discovery reveals prerequisite work (hybrid identity, permission remediation, license gaps), we deliver the assessment portion and pause the agent build until prerequisites are resolved. The build credit carries forward at the original price. Your file server is untouched either way.