PermAffai

AI Readiness

Find out what your organization is ready for.

Two engagements. Two to six weeks. The shorter one gets a policy adopted. The longer one does that too, and maps the work itself against the hours it costs you now.

$2,500 or $5,700, fixed. Scope and price are in the proposal before anything starts.

01 Why this exists

Your staff are already using AI. The question is whether anyone has decided how.

Somebody on the team has pasted part of a homeowner file into a chatbot to get a letter drafted faster. That is not a failure of judgment; it is what happens when the tools are free, the work is piling up and nothing in writing says what is allowed. The risk is not that staff are careless. It is that the organization has not decided anything, so every person decides for themselves.

The second problem is quieter. Nobody has looked at the work itself — which tasks recur, how many hours they take, and which of them a tool could carry without anyone losing control of the outcome. Without that, buying software is guessing.

This is what an AI Readiness engagement settles. Both engagements end with an acceptable-use policy and a data-handling checklist adapted to your organization and ready for your board to adopt. The longer one does that and inventories your own workflows as well, recording what each one costs you in staff hours, so the next decision is made on evidence rather than on a vendor’s demo.

The goal underneath both is plain enough to say out loud: fewer staff hours spent on the work that repeats, without anyone losing control of the result. Our published hypothesis is a reduction of 20% or more in staff hours on targeted routine stewardship workflows within 180 days of a Stuard Companion go-live, measured against each organization’s own baseline — and we have committed to reporting the measured result whether it meets that or falls short.

An assessment is where that baseline comes from, and the timing is the whole point: it is taken before any policy, procedure or tool changes the work. A baseline measured afterwards is not a baseline. So an assessment will not tell you how many hours you are going to save before it has counted the hours you spend now. It counts them, and sets your target against your own figures, with the method written down so you can check it a year later. The hypothesis, the five metrics and the reasoning →

02 Two ways in

Both are fixed-price. Neither requires you to buy anything afterwards.

AI Readiness Consultation

$2,500

2 to 3 weeks

A scoped conversation with leadership, and a short written recommendation.

For organizations that want to make a decision about AI without committing to a full assessment first.

  • A working session with your leadership team on where AI fits and where it does not
  • A policy adoption session — we sit with leadership and adapt an acceptable-use policy and a data-handling checklist to your organization, line by line, so what goes to your board is yours rather than a template
  • A short written recommendation memo, naming what to do next and what to leave alone

Who takes this. An organization that needs a position on AI its board can adopt before staff go any further, and wants that settled without committing to a full assessment.

AI Readiness Assessment

$5,700

4 to 6 weeks · the standard engagement

The full picture: your workflows, your data, your people, and a ranked set of recommendations.

This is what most organizations buy, and what the rest of this page describes.

  • Stakeholder interviews across the organization
  • A workflow inventory in one or two functional areas, with baseline staff hours recorded for each, and a reduction target set against that baseline as a percentage
  • A policy adoption activity — the same acceptable-use policy and data-handling checklist, adapted with your leadership and taken to your board. Named here rather than assumed, because it is the deliverable most likely to be quietly dropped
  • An inventory of the tools you already run, and what would have to connect to what
  • Eligibility assessment for Anthropic’s Claude for Nonprofits program
  • A written assessment report with a prioritized recommendation set and an implementation roadmap
  • A presentation of the findings to your leadership team

Who takes this. An organization ready to act on the answer, and anyone who needs a document a board or a funder will accept as the basis for a decision. It carries the policy work as well, and adds the evidence underneath it: who does what, and how long it takes.

Both prices are fixed and quoted in the proposal. There is no per-seat charge, no software license inside either engagement, and no obligation to proceed to anything afterwards. If the honest recommendation is that you should not adopt AI yet, that is what the report will say.

03 How it runs

What it asks of you.

Five stages, in order. The Consultation is the same shape built around the policy session, without the interviews and the workflow inventory.

Scope

Kickoff and scope confirmation. We ask for your organizational documents and a list of the tools you already use, and work out together who should be talked to. Your time: one meeting, and someone to pull documents.

Conversations

Interviews with the people who actually do the work. Workflow documentation in the areas we agreed, with the hours each one takes recorded as the baseline. Your time: about an hour per person.

The policy session

Leadership reads a draft, we work through it together and adapt it to how your organization actually operates, and you take the result to your board. Your time: one working session.

Analysis

We evaluate each workflow against how often it recurs, how predictable it is, and how much manual effort it carries. Priorities are ranked and the report is drafted, and your champion reviews it before it is final. Your time: one review conversation.

Handover

Final report delivered. Presentation to your leadership team. Handoff documentation so the work is usable without us. Your time: one meeting.

Two to six weeks is typical, but the schedule follows your calendar rather than ours. If a board date, a budget cycle or a short-staffed month sets the pace, say so at kickoff and the plan is built around it. Almost all of what we need is time with people who already know the answers; nobody has to learn a new system to take part.

04 How we work

Three commitments, in every engagement.

These are the questions a board asks, so they are answered here rather than buried in a contract.

  1. People decide. Staff review and approve every notice, checklist and homeowner message a tool drafts, before it goes out. Nothing we build sends anything on its own.
  2. Formulas stay fixed. Where a number matters — a resale price, an affordability figure — it comes from tested, version-controlled code that matches your adopted documents, not from a model’s own arithmetic. The result can be reproduced and audited.
  3. Data stays home. Your homeowner records stay in your own systems. You may choose to share de-identified program data with us; if you do, it is reported to funders and to the field only as combined totals that never identify a single organization.

The policy comes first, not last

A policy adopted before anyone buys anything is the cheapest protection available, and it is the thing most organizations skip. So it is named in both engagements rather than assumed into either: an acceptable-use policy and a data-handling checklist, adapted with your leadership in a working session and ready for your board to adopt.

The Consultation is built around that session. The Assessment carries it too, alongside the workflow evidence that shows where a policy would actually bite. The harder half of the question — how your data is handled today, by whom, and what would have to change before a tool touches it — belongs with a deployment, so it sits in the larger engagement rather than ahead of a decision.

05 Beyond the assessment

If you are ready to build, we are too.

Some organizations already know what they want. They have the policy question settled, they know which work is eating their week, and what they need is the thing itself rather than a report recommending it.

One larger engagement exists for that. Everything in the Assessment is inside it — the same readiness work, the same workflow and protocol mapping, the same policy adoption, measured against the same baseline. It does not skip the diagnosis to get to the build.

Scoped and quoted

Comprehensive Assessment with AI Tool (Stuard Companion) deployment

The comprehensive version of the Assessment first: broader coverage across more of the organization, deeper engagement with more stakeholders, a full adoption roadmap, and an executive presentation built for a board or a funder. For larger portfolios, multi-organization planning, and anywhere the output has to carry weight outside the building.

It is also where the data governance and privacy assessment sits — a proper review of how resident and homebuyer information is held, who can reach it, and what would have to change before a tool touches any of it. That work belongs with a deployment rather than ahead of a decision, because it is only worth doing in detail once something is actually going to be built.

Then, on the same timeline and the same contract, a configured Stuard Companion deployment — your formula, your procedures and your compliance calendar, standing up as something your staff can actually ask. One procurement instead of two.

If you want the comprehensive assessment and are not ready to commit to a deployment, that is how it gets scoped. The assessment is not conditional on buying the tool.

It is scoped and quoted rather than listed at a fixed price, because what it costs depends on how much of your organization is in scope and how much gets built. Ask for a proposal and you will have a number and a scope in writing before anything begins.

06 Who does the work

The person doing the assessment has run the programs being assessed.

PermAffai is a housing practice that works in AI, not an AI practice that found housing. The work is led by Matthew Wyman, who has administered a county down payment assistance program, supported the establishment of a community land trust from inside local government, and built the stewardship and compliance systems this assessment looks for.

That matters for one practical reason: the interviews go faster and the recommendations land closer, because nobody has to explain what a ground lease is, what a resale formula does, or why an expiring restriction schedule is hard to produce.

More about the firm and the people →

07 Request a proposal

A scope and a price, in writing.

Tell us enough to scope it and you will get a proposal naming the engagement, what it covers, what it costs and how long it takes. No obligation attached, and nothing happens until you sign something.

If you would rather talk first, book a call — thirty minutes, no preparation needed.

Everything below except your name, organization and email is optional. The more you fill in, the closer the first draft will be.

We reply within five business days. See the privacy policy for what we do with this.