M.A.I. Consulting
Who we work with

Four roles carry this question. Which one is yours?

Mission-driven and international organisations face the most acute version of the AI question: real pressure to work faster, and real consequences if a beneficiary's data ends up in the wrong place. Below is who we speak with, what each is trying to achieve, and the result each one ends with.

Who we speak with

Four roles, four questions.

Across NGOs, foundations, UN agencies and public institutions, from small teams to 500 staff.

The executive director

Has the board asked for a position on AI, and do you need one before the next meeting?

The job: Answer the board and the funders on AI with a written position.

The problem: A board member has raised the question. A donor questionnaire includes an AI section. Staff are believed to use AI, and the extent is unknown.

The result: A scored baseline and a policy in force before the next board meeting, and a written answer for the donor questionnaire.

Current response: The question is deferred to the next meeting.

How we get there: AI Readiness Assessment →

The data protection officer

Have you found organisational documents in a consumer chatbot, or do you suspect you would?

The job: Prevent the incident, pass the audit, keep the tool list under control.

The problem: A prohibition memo was issued. Use continued out of view, and can no longer be monitored.

The result: A policy people follow because it permits. An approved tool list, a data classification table, an evidence file.

Current response: Prohibition and blocking, with limited visibility.

How we get there: AI Use Policy →

The HR or learning lead

Did you buy a learning platform that most staff never finished?

The job: Deliver training people complete and use, and spend the training line defensibly.

The problem: Completion rates below twenty percent. No evidence of changed practice. Staff still ask what is permitted.

The result: Ninety minutes per department that fits a real calendar. An attendance record and a before-and-after score.

Current response: The licence is renewed.

How we get there: Team AI Training →

The head of programmes or fundraising

Is each proposal drafted with AI in a different way, by colleagues who have not compared methods?

The job: Deliver proposals and donor reports on time without exhausting the team.

The problem: Uneven, private AI use. The risk that a junior colleague sends an unverified figure to a donor.

The result: Countable hours recovered on the next three proposals. A playbook the team uses. Permission to share what works.

Current response: Additional hours and personal accounts.

How we get there: Team AI Training and Custom Tools →

Who decides, who pays, who benefits. The head of programmes feels this most and signs least. The executive director decides and usually pays; the data protection officer decides on policy and can block or sponsor; HR pays from the training line when the trigger is the training cycle. We speak first with the director or the DPO, deliver for programmes, and give HR the evidence it needs.

Training centres and academies work with us as delivery partners. Private-sector work is taken on referral. Individuals and organisation representatives are served through courses for fewer than ten participants, in person or online, priced per person.

Foundations and donors are a separate conversation. If you fund organisations rather than run one, the work is a portfolio programme: one assessment across the cohort, a shared policy each grantee adapts, cohort training, and one report to you. See portfolio programmes →
How engagements are packaged

Five services, bought singly or as one of three ready-made scopes.

Fixed price, fixed scope, agreed before work begins. Build your own scope in the estimator to see an indicative range.

PackageStarting pointResultHowTimePrice
Foundation Up to 50 staff, no first step taken yet A baseline score and one shared page everybody works from. You know what to fix next and can show a funder you have begun. Short readiness assessment, one organisation-wide session, one shared playbook. ~3 weeks Estimate →
Adoption
Recommended
50–500 staff, AI in use across departments with no policy and no shared method A policy in force, four departments working from their own pages, and a before-and-after score. The file a donor or auditor asks for exists. Full ten-dimension assessment, board-ready AI use policy, organisation-wide session, four department labs with playbooks, re-score at 90 days. ~8 weeks Estimate →
Institution Regulated, audited or multi-site, with compliance to demonstrate A documented file mapped to ISO/IEC 42001, the NIST AI RMF and Article 4, and trainers in-house who can run the sessions without us. Everything in Adoption, plus a governance gap plan against ISO/IEC 42001 and the NIST AI RMF, an EU AI Act Article 4 evidence file, train-the-trainer, and twelve months of support. ~12 weeks
+ 12 months
On a call →
How we work with you

Four steps, and you own what remains.

Step 01

Discovery call

Thirty minutes on what is happening with AI in your organisation and which of the five services applies, if any.

Step 02

Baseline

The assessment, plus interviews where the package includes them. You receive a score and a ranked list before anyone is trained.

Step 03

Delivery

Policy drafted together with your people. Sessions built on your teams' real tasks and real documents.

Step 04

Handover

A re-score, the evidence file, and the playbooks in your hands. Engagements close with a capability you own.

Client voices

“The AI Literacy training wasn't just theoretical. It gave our management team the actual tools and confidence to lead our own digital transformation internally.”

BlackBird Training Center

“M.A.I. Consulting provided the intellectual bridge we needed to move our AI initiatives from theory to production. Their methodology is rigorous, transparent, and results-oriented.”

AI for Good Foundation

Disclosure: M.A.I. Consulting's principal also holds a role at the foundation.

Start with the free readiness check, or start with a call.

Both are free. The check takes twelve minutes and you keep the report in either case.