Client Impact

Three engagements. Named, dated, and evidenced.

Most consultancies publish logos. This page publishes what was delivered, on what date, to how many people, and against what standard. Where a client has consented to be named, the client is named. Where consent is still outstanding, the engagement is described and the client is not.

10 of 10SOW learning outcomes delivered, GSU
66Named attendees, single session, Paterson Group
89Rows, English-French AI glossary, federal
2Official languages, one contract, one day
Engagement 01

Government Services Union, Public Service Alliance of Canada

STRATEGIC AI GOVERNANCE AND AI LITERACY · OTTAWA · 23 APRIL 2026

A national union with a member network across the federal public service needed its leadership to understand artificial intelligence governance obligations before its members encountered them at work.

Delivered as a full day in two modules. Module 1 ran 08:30 to 12:00, Module 2 from 12:45 to 16:15, with the National President in the room for Module 2. Content was anchored to the Treasury Board Secretariat Directive on Automated Decision-Making, the FASTER principles, the Algorithmic Impact Assessment (AIA), and Policy Implementation Notice BT48-53/2025. Microsoft Copilot was the primary hands-on platform, set against a comparative matrix covering ChatGPT, Claude and Gemini. Materials were produced to the client's own brand guidelines and handled under Protected B data classification.

Ten of ten Statement of Work learning outcomes were delivered.

Engagement 02

A federal central agency

BILINGUAL AI WORKSHOP, BOTH OFFICIAL LANGUAGES · OTTAWA · 3 DECEMBER 2025

A central agency of the Government of Canada required an artificial intelligence workshop delivered to a federal public service audience in both official languages, on the same contract, on the same day.

AltNexus was prime contractor on both language streams. The French stream was delivered by a subcontracted francophone instructor under AltNexus contract. The engagement produced a reusable asset that outlives it: an 89-row English-French artificial intelligence terminology glossary, verified against TERMIUM Plus, the Government of Canada's terminology and linguistic data bank.

Two workshops. One contract. Two official languages. Every French-language programme since starts from that glossary rather than from zero.

Engagement 03

Paterson Group Inc.

AI USE-CASE IDENTIFICATION IN A LICENSED ENGINEERING PRACTICE · OTTAWA · 9 MAY 2025

A geotechnical and environmental engineering firm wanted to know where artificial intelligence could take real work off its professional staff, not where it sounded promising.

Four hours on site with 66 named attendees, including the sitting President and the Chief Financial Officer. The session covered an artificial intelligence overview and prompt engineering, then ran a facilitated use-case identification workstream. The applied work went where the firm's risk actually sits: interpretation of the Ontario Building Code and Ontario Regulation hydrogeological guidelines, using the AltNexus CREATE prompt framework. The resulting activity deck was produced in the client's own branding and reviewed by a client subject-matter expert before use.

The session ended with four numbered use cases, each with a named owner, and an AI Champions roster to carry them.

David Gilbert, P.Eng., Past President and Division Director, Geotechnical, has consented to act as a reference.

The pattern

What these three have in common.

Each engagement ended at a decision the client owned: a governance obligation understood, a terminology standard established, or four use cases with names against them. None ended at a report.

[GOVERNANCE]

Governance before deployment

Obligations are named against the instrument that creates them, the Directive on Automated Decision-Making and the Algorithmic Impact Assessment, before any tool is selected.

[CAPABILITY]

Capability that compounds

Every engagement leaves a durable asset behind, a glossary, a champions roster, a scored use-case list, so the next piece of work starts further along than the last.

[DECISION]

A decision, not a deliverable

Work is sequenced to end at a scale, refine, or stop decision the client can defend, with a named owner against each item and a date.

What is not on this page.

No client has yet published a measured dollar outcome from an AltNexus engagement, so none is claimed here. That measurement begins at the next engagement, where the cost of the problem is captured in the client's own numbers before any work starts, with 30-day and 60-to-90-day reviews booked at signature. When those figures exist and the client consents to publication, they will appear on this page.

Book a 30-minute discovery conversation See the engagement model

Engagement facts on this page are drawn from executed Statements of Work, signed invoices, and delivered materials held on file. Client names are published only where the client has confirmed consent in writing.