I ship AI systems into production, and tell you when not to build one.
Cognitiv takes AI systems from decision into production, and then gets the organisation to actually use them. Fourteen years across technology and business. Four systems built and delivered.
The stack the systems run on
Decide, adopt, rely. Most AI programmes fail at one of the three.
Nobody adopts the tools, or nobody can say whether the output is good enough to rely on. Cognitiv works on both - and the second is where it is unusual, because the founder spent five years marking inside an externally moderated examination system before technology.
Decide
Where AI belongs in the business, where it does not, and what has to be true before it ships. Build, buy, or do nothing - in writing.
Adopt
Briefings for the people who will use the tools, and adoption sprints that get one tool into daily use and measure whether it stuck.
Rely
The rubric and human review that stops bad AI output shipping, and ongoing product direction where a wrong output has a cost.
Five engagements. Fixed scope, quoted up front.
Each one starts with a call, is scoped and quoted in writing, and runs under a written agreement. The scope decides the price, so no retainers are published.
AI Readiness and Risk Review
Where AI belongs in the business, where it does not, and what has to be true before it ships. Ends with a written decision, including build, buy or do nothing.
For · Owners and leadership teams deciding whether and where to use AI
Executive and Team Briefing
Practical AI use for leadership teams and staff, built from systems Cognitiv runs in production, not slides.
For · Boards, leadership teams and the staff who will use the tools
AI Adoption Sprint
Gets one tool into daily use. Pick the workflow, set the rules, train the people, measure whether it is actually being used.
For · Teams that bought or built something and nobody uses it
AI Output Quality Gate
The rubric and review process that stops bad AI output shipping. Defines good enough, builds the rubric, stands up the human review.
For · Anyone putting AI output in front of customers, regulators or students
Product Direction
Ongoing, month to month, for AI-dependent products in compliance-sensitive domains: architecture, policy constraints, release management, data-protection posture.
For · Products where a wrong output has a cost
Not sure which fits
Start with a call. It covers what you are trying to do, which engagement fits, and whether Cognitiv is right for it. If it is not, that gets said on the call.
Written agreement before any work starts. See the full detail of each engagement →
Built and delivered.
AI scoring and evaluation platform
Specified and shipped from zero. Production SaaS at scorafy.com in two regions - rubric engine, human release gate on every automated result, Stripe billing, configurable PII retention. Cognitiv-owned. In production use with a design partner.
Digital platform and member hub
Technology Partner to Sensational League from December 2025 to September 2026. Built and led the digital platform and member hub - member profiles, GDPR-compliant magic-link sign-in, and the match-day check-in and goal capture used at the first live event.
I'm Adam Broons.
CTO at OopZee, a Danish-built vehicle-damage reporting app released on the Apple App Store and Google Play. Founder of Scorafy - an AI assessment platform running in two regions, in production use with a design partner. Before technology, five years marking inside Western Australia's externally moderated examination system.
Background - ANZ small-business specialist, three years · secondary computing, Department of Education WA, five years · Vistage AI speaker, two CEO-group sessions · capital-raise documentation for Pentanet ahead of its IPO.
More about how I workWhat this is, and what it costs.
What does an AI delivery consultant do?+
Takes an AI decision through to a system running in production and in daily use. Cognitiv sets the architecture and the rules, directs the build, reviews and rejects output, and owns the result. It also says when not to build - the work Adam is proudest of is a rostering system he talked a client out of.
What does it cost?+
Every engagement is fixed scope and quoted up front after a first call, with a written agreement before any work starts. No retainers are published because the scope decides the price.
Who is it for?+
Owners and leadership teams adopting AI, founders shipping AI-dependent products in compliance-sensitive domains, and any organisation that needs to know whether its AI output is good enough to rely on.
Do you write the code?+
No. Cognitiv directs AI development tools and is accountable for what ships: the architecture, the policy constraints, the review, and the result in production. It does not hand-write the implementation and does not claim depth it does not have.
Deciding where AI belongs in your business?
Start with a call. Fixed scope, quoted up front, and a straight answer if the timing is not right.