According to a recent LinkedIn post from Sifflet, discussion from the company’s Signals25 summit suggests that many large enterprises are experimenting with AI without yet demonstrating clear, measurable impact on profit and loss statements. The post references perspectives aligned with commentary from McKinsey and MIT, and highlights speaker Thomas Krakty’s view that most deployments struggle in practice.
Claim 55% Off TipRanks
- Unlock hedge fund-level data and powerful investing tools for smarter, sharper decisions
- Discover top-performing stock ideas and upgrade to a portfolio of market leaders with Smart Investor Picks
The company’s LinkedIn post highlights a key barrier as the lack of “context” around enterprise data, distinguishing this from technical interfaces such as the Model Context Protocol (MCP), which the post describes as an API rather than a framework for meaning. According to the post, challenges arise when AI agents are unable to reconcile how different business functions define core metrics or identify authoritative data sources.
The post suggests that this contextual gap is where many enterprise AI initiatives fail quietly, implying that successful AI orchestration may depend as much on data governance and semantic alignment as on model choice or infrastructure. For investors, this framing points to a potential demand area for tools that operationalize business context for AI, a segment in which Sifflet appears to be positioning itself.
If the company can translate this narrative into concrete products that help customers achieve measurable P&L impact from AI, it could improve its value proposition in a crowded data and AI tooling market. More broadly, the emphasis on context over pure connectivity may signal an industry shift toward higher-value “intelligence” layers that sit between raw data systems and AI agents, which could benefit vendors able to prove ROI in this niche.