Why I built Intent Solved: Productivity for the 98%

Most of the AI content I read isn't written for the people who do the work

Bottom Line: The 98% of operators running real Australian businesses deserve production-grade AI engineering, not pilot theatre. In 20 years across four tech waves, I've seen almost no one build for them. That's why I built Intent Solved.

The Context

I've spent 20 years delivering emerging tech into Australian enterprises. ERP, then integration middleware, then cyber security, now AI. I've watched every one of those waves land the same way, and I've lost patience with the pattern.

Every wave splits the market into two audiences. A small vocal group of early adopters, consultants, and influencers who ship opinions faster than they ship code. And a much larger group, the 98%, who have a P&L to run, a team to keep shipping, and a board that wants an answer on the new technology by next quarter. The first group gets the content, the conferences, and the attention. The second group gets the bill.

Why this matters: The 98% carry the business. If no one engineers for them, the wave doesn't deliver. Four waves in, I've decided that's the job.

The Insight

The movie I've seen three times already

When ERP landed in Australian mid-market enterprises, half the deployments I saw became a decade of rework. Bought before the business process was understood. When integration middleware arrived, I cleaned up the fallout for another five years. Point-to-point sprawl dressed as architecture. Cyber security went the same way in my experience. Bolt-on controls instead of engineered governance. Audit findings instead of production design.

Every one of those waves started with the same promise and ended with the same unpaid debt: the tool was bought before the problem was understood, and the 98% were left carrying the cost.

AI is tracking the same arc. I've sat in boardrooms where the AI strategy is a vendor shortlist. I've walked engineering floors where "the AI pilot" is a Slackbot and a README. The content being published tells the 2% what to think about agent architectures. The 98% need someone to tell them what to build, in what order, and at what risk.

The 98% aren't behind. They're being served by the wrong end of the market.

The early-adopter 2% run experiments. The 98% run businesses. Those are different jobs, with different risk profiles and different outcomes.

A client I worked with earlier this year had 16 developers, a live product shipping to paying customers, and a board asking how they were using AI. The question wasn't "which agent framework should we pilot." The question was "how do we deploy Claude Code without burning two weeks of velocity and without causing an incident." That's an engineering question, not an opinion-piece question. Most of what they found online answered neither.

This is the pattern I see in almost every engagement. Capable engineering teams. Clear mandates from the top. No production outcome, because the market sold them experimentation when they needed engineering.

Productivity for the 98% looks different from productivity for the 2%

For the 2%, productivity is new frameworks per quarter, tokens per minute, and impressions per cycle. For the 98%, productivity is measurable return on the AI investment the board already approved. Return on Intent, not return on hype.

That's the target I build for. Governance the team can actually run. Playbooks with the hard decisions already made. Claude Code advisory that gets enterprise engineering teams shipping safely inside four to six weeks. When I tell a client AI is not the right answer for a particular problem, I mean it. The 98% notice when someone isn't selling them the default.

What This Means for Your Business

The pattern: The AI market is selling experimentation. The 98% need engineering. Those aren't the same service, and pretending otherwise is how pilot purgatory happens.

How to Act on This

Immediate Actions (This Week)

1. Audit your AI content diet. If 90% of what you're reading is written for the 2%, you are spending attention on the wrong signal. Replace it.
2. Name the production outcome you need. "We use AI" is not an outcome. "We cut Tier 1 support cycle time by 30% with a governed deployment" is.
3. Apply the problem taxonomy before the tool. People, process, data, integration, technology, then AI. Every time, in that order.

Strategic Actions (This Quarter)

• Stop buying experimentation. Start buying engineering. Scope engagements in outcomes, not hours. Ask the provider to define what "shipped" looks like before you sign.
• Build internal capability, don't lease it. The 98% who get return on their AI spend are the ones who own the operating model, not the ones renting a vendor's.

Framework Application

The .solved Execution Framework is built for the 98%. Step 1 (Uncover) surfaces the real problem, not the presenting symptom. Step 2 (Unpack) checks people, process, and data before anyone touches AI technology. By Step 4 (Embed), we are shipping production-grade capability, not another pilot that quietly dies at the six-month review. The framework exists because the 98% can't afford theatre.

The Bottom Line

I built Intent Solved because the 98% are the ones keeping Australian enterprises running, and almost no one in the AI market is engineering for them. Twenty years across ERP, integration, and cyber taught me two things. The waves repeat. The quiet majority always pays for the noise the loud minority makes. I'm not interested in running that pattern a fourth time from the stands.

If you're in the 98%, you are the business. This practice is built for you.

Next Steps: If you are carrying an AI mandate and need to ship production outcomes, not pilot summaries, get in touch. claudecode.solved Advisory takes enterprise engineering teams from pilot purgatory to governed velocity in four to six weeks.

Related Resources

• The .solved Execution Framework - Our five-step framework for moving from pilot purgatory to production-grade AI capability
• claudecode.solved Advisory - Our service for enterprise engineering teams deploying Claude Code into production.
• Hard Hat Era, not Experimentation Era - Why the next phase of AI adoption is engineering, not exploration.