- Published on
Creating AI value with enterprise context
- Authors
- Name
- DeepHorus
Only 16% of organizations report extracting transformative value from AI, according to a recent MIT Technology Review Insights poll. The survey shows most companies still struggle to realize AI's full potential. Many see only moderate or minimal returns.
Businesses deploy various AI tools, including customer platforms, conversational assistants, and workflow automation. These systems are sophisticated but often lack proper understanding of the business context. This can lead to outcomes that are wrong, confused, or misaligned with goals.
The poll reveals 40% see moderate gains in specific tasks or teams. Another 21% report meaningful gains, while 23% experience minimal or no returns. Most companies are still piloting AI (38%) or using it to assist employees (35%). Only 12% have AI agents in production, and just 3% use AI for most work in key processes.
The missing layer is context, according to Prukalpa Sankar, co-founder of Atlan. A context layer sits between business systems and AI platforms. It translates enterprise intelligence into governed, machine-usable context for AI agents or models. This helps knowledge compound.
Sankar said without a context layer, AI is like a blindfolded smart intern. "Itâs almost like youâre hiring the smartest intern in the world, blindfolding them, and throwing them into your organization," she said. She noted that while artificial general intelligence is already here for every real-life intent, itâs not yet useful in most companies.
Establishing a context layerâor a "company brain"âcould enable more accurate, cost-effective AI. It could support autonomous agents and help organizations move beyond pilots. Investing in enterprise context may be the key to transformative value.
