In a wide-ranging discussion moderated by Twilio CMO Chris Kohler, engineering leaders from Zillow and Nextdoor revealed how they are driving AI integration across customer engagement, internal productivity, and platform development. The session was informal and candid, touching on organizational challenges as much as technical strategies.
Joel Kapelyov, Director of Engineering at Zillow, and Krish Sailam, Head of MarTech and AI Ops from Nextdoor, explained how AI is reshaping their operations. Both organizations are focusing on building and evolving tech stacks that support multi-persona experiences, developer enablement, and customer-level personalization.

Zillow: Building a Composable, AI-Native Stack
Zillow’s platform handles lead ingestion, outreach, and handoff across agents, customers, and loan officers. Built heavily on Twilio for communications, it now includes a centralized AI team supporting model development, prompt engineering, and compliance with fair housing regulations.
Zillow’s approach mirrors Twilio’s ethos: intuitive APIs, modular architecture, and problem-first design. Custom models and LLMs are embedded into platforms like Cursor AI, while acquired tools use AI to summarize interactions and automate agent workflows. Despite AI integration, human agents remain central to negotiation and advocacy.
Zillow also shared early thinking about future agentic interfaces, including unified customer profiles supported by multiple AI agents for different tasks, from neighborhood recommendations to loan eligibility screening.
Nextdoor: From Buying Tools to Building AI Systems
Krish described how Nextdoor is evolving from a buy-first mindset to building bespoke tools that span MarTech, AI operations, and customer service. AI is used internally to speed up sales and legal processes, and externally to enhance user experiences with features like AI-generated classifieds.
Rather than build individual AI stacks for each function, Nextdoor is creating a shared AI operations layer that supports sales, marketing, customer support, and internal teams. Executive support and cross-functional collaboration have been key to rapid experimentation and adoption.
Internal hackathons are now fully AI-focused, and developers are encouraged to explore use cases hands-on. AI is also embedded into software development practices, PRD creation, and documentation workflows, reducing cycle time and boosting velocity.
Communication and Customer Readiness for AI
Both Zillow and Nextdoor emphasized that successful AI integration depends on communication. Whether it is an agent using AI-summarized messages or a product manager leveraging AI to draft a roadmap, the expectation is that AI enhances context rather than replaces human input.
Kapelyov noted that customers are more open to AI agents than expected so long as the experience feels trustworthy and contextually aware. AI agents, when branded well and backed by human escalation paths, can build customer confidence without degrading quality.
From CRM integration to message orchestration, Twilio serves as the connective tissue in these efforts. Its APIs support the low-latency, multi-channel communication required to bring these AI-powered workflows to life.
Personal Agents and the New Interface Layer
The session closed with a forward-looking discussion on how AI may fundamentally change user behavior. Twilio, Zillow, and Nextdoor see the potential for AI to replace traditional websites and apps with personal agents that handle tasks via natural language prompts.
These agents would tap into unified customer data and integrate with messaging platforms to perform actions without forcing users to navigate static interfaces. The shift to an agent-first interaction model could reshape product strategy, data governance, and competitive differentiation.
As AI reshapes the way consumers engage, Twilio’s platform continues to provide the essential rails for data-rich, real-time communication that adapts to these evolving patterns.

