{
  "id": 6428531,
  "title": "From AI pilots to production: Rebuilding the enterprise stack for intelligent agents",
  "url": "https://urgent.news/2026/09/09/from-ai-pilots-to-production-rebuilding-the-enterprise-stack-for",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-09T13:27:58.000Z",
  "source": {
    "name": "YourStory",
    "slug": "yourstory",
    "url": "https://yourstory.com/2026/09/ai-pilots-production-rebuilding-enterprise-stack-intelligent-agents"
  },
  "original_language": "en",
  "account": "Enterprise AI discussions have shifted from model selection to the broader equation, a central theme at DevSparks Hyderabad 2026. Kiran Rokkam, Partner AI/ML at Tiger Analytics, and Naren Peri, VP of Data & Analytics at MetLife, joined moderator Shivani Muthanna for a discussion titled 'The new enterprise stack: Data, agents, decision intelligence'.\n\nRokkam noted that enterprises are moving from AI exploration to dedicated transformation functions, with about 80% of their clients having transformation practices. AI is increasingly tied to business operations rather than remaining an experimental area. Peri emphasized that AI is no longer limited to specialist teams and that everyone can benefit from frontier models, but with risks that must be acknowledged.\n\nMoving from copilots to agents, data quality and context are crucial. Enterprises need AI-ready data, which involves more than just cleaning databases; they must make their context understandable to machines through ontologies, knowledge graphs, and semantic layers. Rokkam highlighted that much of an enterprise's knowledge doesn't reside in structured databases but in various formats like emails, presentations, SharePoint, Google Drive, and team conversations. The challenge lies in understanding business processes and employee knowledge, as legacy infrastructure often has dispersed employee, customer, or operational data using different identifiers and definitions. The first step, according to Rokkam, is creating a context layer with knowledge graphs in each system.\n\nThe panel refuted the notion that enterprise AI is about replacing human work, emphasizing that machine scale combined with human judgment is the immediate opportunity. In industries like insurance, decisions significantly impact people's lives, requiring a balance between machine speed and human empathy and moral judgment. Finally, Rokkam stressed that productivity is a key metric, but enterprises are increasingly focusing on how AI translates into actual business value.",
  "summary": "At DevSparks Hyderabad 2026, leaders from MetLife and Tiger Analytics explored what it will take to move enterprise AI from experimentation to production, from AI-ready data and semantic layers to governance, human judgment, and measurable business outcomes.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}