The foundational elements of AI architecture that IT leaders need to scale
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Elastic sponsored the piece through MIT Technology Review's Insights arm, positioning data quality, context engineering, governance, and human expertise as the four architectural elements that will outlast shifting model capabilities
- Gartner predicts 60% of AI projects will be abandoned through 2026 if they are not supported by AI-ready data, citing inconsistent structures, fragmented ownership, and incomplete datasets as the main culprits
- Adnan Adil, CIO of Elastic, calls data "a durable part of AI architecture" and warns that AI itself cannot fix underlying data problems, meaning pipelines, labeling, and ownership must be designed in from the start
- Context engineering — distinguished from prompt engineering — relies on RAG, vector databases, and careful prioritization, because feeding models too much context dilutes relevance, drives up token costs, and slows responses
- An Elastic 2026 report found 85% of IT decision makers expect to enable LLM observability for their internal generative AI apps, with Adil calling observability "huge" for cost control, decision-making, and engineering efficiency
- Deloitte's 2025 Tech Executive Survey shows nearly 70% of tech executives plan to grow their teams in response to generative AI, a figure the piece highlights as a counterweight to widely reported AI-related layoffs
- AI governance is framed as inseparable from security: the piece flags prompt-based data leakage, model vulnerabilities, and adversarial inputs as expanded attack surfaces that require access controls embedded from day one
Why it matters: IT leaders face a Gartner-projected 60% AI project abandonment rate by 2026 if data foundations aren't built first — meaning the choice Elastic pushes isn't between vendors but between architectures designed to scale agentic systems versus ones destined to be scrapped. The Deloitte finding that 70% of tech executives plan to grow teams in response to generative AI directly contradicts the layoff narrative and puts human expertise on equal footing with technical infrastructure.
