Becoming an AI‑native enterprise means rethinking how the business operates so that AI agents are embedded in day‑to‑day work, not just added as one more tool.

From the conference discussions, AI‑native organizations:

• Use AI agents to do tasks, not just provide insights. As Arvind Jain from Glean put it, agents are about “doing the tasks” and effectively coworking with humans, especially in labor‑intensive areas like customer experience and more deterministic work like coding.
• Redesign business models and workflows around AI. Jai Das from Sapphire Ventures emphasized that companies that succeed will change their business model, code development practices, and go‑to‑market motions to take advantage of agents.
• Invest in leadership and governance. Strong CEO and management sponsorship is described as essential. Leaders must be “fully behind the effort” in how AI is developed, sold, and managed across the company.
• Build on top of models and agents. Enterprises are still in the “early innings” of learning how to create the right harnesses and guardrails so agents can reliably perform work based on underlying AI models.

In short, AI‑native enterprises reimagine their operations so AI agents become core coworkers in financial analysis, customer experience, and coding, rather than side projects or isolated pilots.

Enterprises are starting to use AI agents to proactively understand and address customer needs, not just to deflect calls.

A concrete example is T‑Mobile’s AI‑native push:

• Proactive intent detection. Working with OpenAI and Distyl AI, T‑Mobile built an application called IntentCX to interpret customer “breadcrumbs” — signals about issues or reasons they might leave — and solve problems before customers feel the pain.
• High coverage of AI‑handled interactions. T‑Mobile has deployed bots on its app and website, plus voice bots in the app and on the phone channel. According to Julianne Roberson, voice AI now answers more than half of all phone calls.
• Focus on experience, not just cost. The goal is to remove unnecessary calls from customer care by delivering a high Net Promoter Score (NPS) experience and resolving issues early, rather than forcing customers away from human agents.
• Leadership engagement. Roberson highlighted that when roadblocks appear, senior leaders — including the CIO and CEO — are directly involved in working through them. This level of engagement is seen as critical to making AI‑native customer experience work.

The early pattern is clear: AI agents are being used to reimagine customer journeys, with measurable shifts such as a majority of calls now being handled initially by AI, while still aiming to improve satisfaction rather than simply cut support costs.

AI coding agents are starting to reshape how software is built and maintained, with implications for both internal development teams and SaaS providers.

From UiPath’s perspective, shared by CTO and CPO Raghu Malpani:

• Agents as builders and managers of the SDLC. UiPath is “going all in on coding agents,” aiming to expose its platform as APIs and CPIs so agents can automate large parts of the software development lifecycle (SDLC), from building to managing processes.
• From coding to composing. On SaaS platforms, coding agents increasingly act as composers rather than line‑by‑line coders. For example, UiPath provides building blocks, and the agent configures these to achieve a process outcome instead of coding everything from scratch. This can significantly reduce development time.
• Governance remains a key SaaS advantage. Even if agents can build software from the ground up, SaaS vendors still provide governance capabilities such as auditability, traceability, compliance, and certifications — all of which are critical for enterprises.
• Cost and scale considerations. As Jai Das noted, enterprises will need to watch the rising cost of using large models and the tokens required for development. This will influence how aggressively they adopt coding agents at scale.

Overall, coding agents are pushing enterprises to rethink their build vs. buy decisions. Instead of fully replacing SaaS, they are more likely to change how SaaS platforms are used — with agents orchestrating and composing capabilities while vendors continue to provide governance and reliability.