Crema, a design and technology consultancy based in Kansas City, MO, was engaged by a leading energy-dedicated SaaS company to explore and pressure-test the viability of an AI-powered land title chain automation platform, with the goal of informing a strategic build-vs-acquire decision.
The client (IP remaining private) provides data, analytics, and software to oil and gas companies worldwide. With an existing digital land record advantage and deep domain expertise, the company saw an opportunity to expand into a compounding, AI-enabled platform capable of capturing additional markets and unlocking new economic profiles, starting with the historically complex and labor-intensive world of land title chain construction.
Over just three weeks, Crema's team worked against a real-world dataset of land records dating back to 1880, building and testing an automated title chain pipeline that ingested, parsed, and connected historical documents through logic embedded in the source text.
The prototype achieved 44% accuracy with non-expert human assistance, a meaningful proof of concept that also revealed the depth of challenge ahead. Primary challenges were rooted in the inherent messiness of historical records, where misspellings, abbreviations, and inconsistent language proved difficult to resolve at scale.
"The energy data space is riddled with tricky problems, and land title is one of the hardest. What we proved is that AI can dramatically accelerate discovery, but the last mile of confidence still requires human expertise. Understanding exactly where that line sits is the strategic insight."
- Tyler Hilker, Crema
Based on its findings, Crema delivered a clear strategic recommendation, supported by a full feasibility report, functional prototype, and implementation roadmap to help the client move forward with confidence.
