There’s a new term making the rounds in AI circles: “Context Engineering.” When I first heard it, I had to smile — because I’ve been doing it for over 20 years. We just called it Business Analysis.
The name might be new, but the work isn’t. It’s about understanding how systems connect to real-world business needs — and how knowledge that lives in people’s heads gets translated into something machines (and organizations) can actually use.
________________________________________
How Building With AI Changed My Perspective
About a year and a half ago, I started building AI-powered tools for myself — a portfolio tracker, a music library, a collectibles catalog. Nothing fancy. Just projects I wanted.
And every one of them taught me the same lesson: AI doesn’t know what it doesn’t know. I could ask AI to “build me a portfolio tracker,” and it would create something functional. But would it know that I wanted to track the exact cash I put into each account? That the $4,000 I invested in one Roth IRA in 2001 grew to $104K, while the same amount in another barely hit $20K?
Of course not. That context — what matters to me, how I think about my money, and what questions I actually need answered — only exists in my head.
That’s when it clicked: context isn’t data. It’s meaning. And meaning still needs a human.
________________________________________
The Enterprise Parallel
The same thing happens in enterprise systems — just with bigger stakes.
In my work for Disney, Warner Bros, Pokémon, and Amazon MGM Studios, I’ve seen how every organization develops its own internal language. Even when two companies use the same terms — territory, media type, start point, product, approval process — they rarely mean the exact same thing.
One studio’s “territory” might include Canada under North America; another treats it as a separate region. One company’s “media tree” lumps SVOD and AVOD together; another splits them by platform. A “start point” or “waterfall” in participations can follow radically different recoup orders. Even “product approvals” vary — some rely on automated workflows, others on decades-old email chains and tribal sign-offs.
That context isn’t in the documentation. It’s scattered across systems, spreadsheets, contracts, emails, and the analysts who remember why each exception exists.
AI can’t infer that on its own. Someone has to know it exists, know where to find it, and know what matters.
That’s Business Analysis. That’s context engineering.
________________________________________
Why This Matters More Now
Yes, AI can code. I use it every day. But the code is only as good as the requirements behind it. If I tell AI, “build an investment app,” it’ll build something generic. If I want it to handle my transaction types, my logic for cash flow, my edge cases — I have to teach it. Layer by layer. Context by context.
The same applies to enterprise systems.
You can ask AI to design a licensing or participations platform, and it will give you something that looks reasonable. But will it reflect how your company actually calculates participations or defines rights? Will it account for the quirks hidden in five legacy systems or those unspoken “we always do it this way” rules?
Not unless someone who deeply understands the business guides it there.
________________________________________
Context Engineering = Business Analysis + Domain Expertise
The AI world is rediscovering what Business Analysts have always known: Systems only work when they reflect business reality.
Now, we simply have a new partner in the process. AI helps us think through edge cases, stress-test logic, and prototype faster — but it still needs human judgment to define what’s true, what matters, and what the system should care about.
That’s not prompt engineering. That’s not just writing better instructions. That’s context engineering — and we’ve been doing it all along.
________________________________________
The Evolution of the BA Skillset
We’ve always adapted:
- Waterfall → Agile
- Documentation → User Stories
- Spreadsheets → Jira
Now we’re adding AI as a thought partner, brainstorming tool, and rapid-prototyping engine. But the core job hasn’t changed: understand the business, capture the undocumented, and translate it into something systems can use.
If anything, that role is more critical than ever.
________________________________________
The Bottom Line
Until AI can autonomously discover, validate, and prioritize undocumented business knowledge, Business Analysts aren’t being replaced. We’re just getting the most powerful toolkit we’ve ever had.
________________________________________
Ross French is Principal Consultant at Skoopix, specializing in Business Analysis and Project Management for Media & Entertainment. With 25+ years leading enterprise transformations at Disney, Warner Bros, Amazon MGM Studios, and Pokémon, he now applies hands-on AI development to explore how intelligent systems can enhance strategic business thinking.