Insights from COBOL thought leaders and the brightest minds in the business — on AI, semantic reasoning, and the future of enterprise software modernization.
Connect With Our ExpertsAnthropic's latest claims about Claude's COBOL proficiency have the enterprise world buzzing. But there's a profound difference between a language model that has read COBOL and a system that truly understands it.
Anthropic's latest claims about Claude's COBOL proficiency have the enterprise world buzzing. But there's a profound difference between a language model that has read COBOL and a system that truly understands it.
In the Information Technology sector, chaos isn't an aberration; it's the operating system. Discover why Strategic Operating Knowledge is the constant that anchors sound IT strategy when everything else is in flux.
Technology alone rarely determines success. What separates thriving enterprises from failed ones is Strategic Operating Knowledge — the disciplined accumulation and systematic exploitation of how things work in practice.
Your documentation problem isn't as simple to fix as you think. GenAI alone achieves only 74% accuracy — and for regulatory compliance, modernization specs, and incident post-mortems, that 26% error rate is a catastrophe waiting to happen.
You've been there. Maybe during an OCC exam prep meeting, or a Dodd-Frank compliance review. Your compliance team needs answers about business logic that's been accumulating in your codebase for three decades. They need it documented, traceable, and yesterday.
The enterprise software industry has spent decades pursuing a fundamentally flawed premise: that system value lies in the technology itself, not the business logic it contains. Before you write the next check for a modernization project, ask: do you actually understand what you're trying to replace?
220 billion to 800 billion lines of COBOL still power 80% of the world's financial transactions — built by a generation now walking toward the exits. When your experts retire, they're taking the unwritten manual with them.
An estimated 220-800 billion lines of COBOL code remain actively deployed across global enterprises, processing trillions in daily transactions. Yet over 80% of organizations expect these systems will persist beyond their current workforce's tenure.
Research Scientist Aleksander Chakarov, Ph.D., presented a Phase Change workshop paper at the 32nd AAAI Conference on Artificial Intelligence, sharing lessons learned in implementing a natural language chat interface for software engineering.