Step 1 before you
build AI.
We help Ontario businesses and organizations plan, assess and adopt AI safely, so what you build is useful, trusted and ready to use.
Aligned with government AI guidance and privacy requirements. Independent: we don't sell AI tools.
Many are rushing to build AI. Few start with Step 1.
AI is moving from helping to acting
It can now issue refunds, update records and send payments. The question is shifting from what AI can do to what it should be allowed to do.
Building the wrong thing is expensive
Without a clear problem, business case and risk check, organizations invest in tools that solve the wrong problem or create privacy risks.
Most teams have no one to do this
Large companies have AI and risk teams. Most organizations don't, so they either avoid AI and fall behind, or adopt it without guardrails.
Step 1: the AI Readiness & Safety Assessment
Six questions answered before anyone builds or buys an AI solution.
Discovery and needs analysis
What problem are we solving, for whom, and is AI the right fix?
You getA clear problem statement and priorities
Process mapping
How does the work happen today, and where would AI fit?
You getA process map with AI opportunities marked
Business case
What will it save or improve, what will it cost, and how will we measure success?
You getExpected benefits and success measures
AI risk assessment
What could go wrong with data, privacy, errors or bias?
You getA risk register and the safe AI level for each task
Governance and oversight
Who approves, monitors and is accountable for AI decisions?
You getAn AI use policy aligned with government guidance
Build-ready brief
What exactly should be built or bought?
You getRequirements you can hand to any developer or vendor
How we decide what AI may do
The risk assessment places every task at one of four safe AI levels.
AI does it alone
Email drafts, meeting notes, summaries
AI suggests, a person decides
Reports, quotes, recommendations
AI acts only with approval
Refunds, invoices, record changes, signed off and logged
AI never touches it
Payroll, payments, confidential personal data
Guardrails, tools and training
Responsible AI governance
Practical AI use policies, human oversight practices and clear accountability for how AI is used.
AI adoption and workflow improvement
Choosing suitable tools and supporting AI-assisted workflows, reporting, dashboards and automation.
AI literacy and training
Helping leaders and teams understand what AI can and can't do, and how to use it responsibly.
AI use review
Already using AI? We review how it's used today and where the risks and gaps are.
Any organization ready to use AI responsibly
Small and medium businesses
Want the time savings of AI but have no IT or risk team.
Larger organizations and teams
AI is already in use, but without shared rules or oversight.
Not-for-profits
Limited budgets and sensitive personal data to protect.
Professionals and students
Workshops and talks on safe, effective AI use.
Working with AI builders
Our role is to help you decide what to build, why, and how to keep it safe. We don't compete with developers and vendors; we make their work more successful. You come out of Step 1 with clear, build-ready requirements, and builders get clients who know what they need.
Anterpreet is a PMP-certified digital delivery lead with 6+ years leading web, data and AI projects at SkipTheDishes, eShipper and Remitr Payments. She has built AI-powered dashboards and reporting tools, written business cases for enterprise projects, and trained leaders and non-technical teams to adopt new tools. She founded K-Connect to give every organization the Step 1 groundwork that large companies take for granted.
- 6+years leading web, data and AI projects
- 30+business cases written, with risk and issue logs (RAID)
- 400+digital assets brought to WCAG 2.1 AA / AODA standards