August 2025 Summaries

Intercom has undertaken a significant transformation in its frontend engineering approach by transitioning from Ember to React, driven by the need to enhance productivity and adopt modern tools. Initially, Ember was chosen for its strong, opinionated framework, which supported rapid decision-making and problem-solving. However, by 2023, the limitations of the outdated framework, such as lengthy build times and stagnation in adopting new technologies, necessitated a change. The decision to switch to React was based on its existing integration in Intercom's infrastructure, extensive ecosystem, and compatibility with AI-driven tools that are integral to the company's future. This strategic move, dubbed “The Future of Frontend,” is aimed at doubling productivity by reducing friction and leveraging AI to enhance developer efficiency, with React now being the default for new UI developments. The transition has already shown positive results, with faster build times, increased developer satisfaction, and a substantial number of new features being shipped. Intercom sees this shift not just as a change in technology, but as a critical step towards fostering innovation and maintaining high standards in its engineering practices, with an ongoing invitation for engineers to join their ambitious journey.

Implementing an AI Agent fundamentally transformed the team dynamics and roles within the company, as experienced during the launch of Fin, an AI Customer Service Agent. The process involved navigating numerous challenges, including the absence of a clear roadmap, the need for rapid deployment, and the integration of new roles like Conversation Designers and Knowledge Managers. Despite initial uncertainties, the AI Agent significantly improved support operations, resolving over 75% of inbound support volume and leading to the establishment of a new AI Support function. The experience inspired the creation of The AI Agent Blueprint, a resource designed to guide other support teams in effectively deploying and scaling AI technology, reflecting the lessons learned and the evolving landscape of AI in customer service.

AI Agents are rapidly transforming business service delivery, customer loyalty, and value creation, with expectations that they will soon manage most customer service interactions. Human roles in customer service will shift towards performance analysis, system improvement, and customer experience design. The AI Agent Blueprint serves as a strategic guide for leaders in support, customer experience, and AI transformation, aiming to facilitate rapid implementation, confident scaling, and meaningful business change with AI. The Blueprint is divided into two parts: "Launch it," which focuses on unlocking immediate efficiency and building a business case for AI, and "Scale it," which emphasizes the need for intentional system rewiring to fully realize AI's potential. This includes creating AI-first customer journeys, embedding AI into organizational structures, and redefining support measurement and funding to transform it into a strategic growth component.

AI adoption can be challenging for companies, as many teams face common pitfalls that hinder success, such as treating all AI tools the same, assuming they can build AI agents in-house, choosing the wrong vendor, neglecting their internal content foundation, and expecting instant perfection. To effectively evaluate and deploy AI, it is crucial to understand foundational concepts like Retrieval-Augmented Generation, vector search, Agentic AI, and MCP (Model Context Protocol), which can enhance decision-making and vendor evaluation. Companies often fail by focusing on flashy demos and outdated technology rather than partnering with innovative vendors who invest in AI R&D and adapt to changing workflows. Additionally, high-quality content is essential for AI systems to function optimally, and setting realistic expectations for resolution rates can lead to improved outcomes over time. A strategic approach, as outlined in "The AI Agent Blueprint," can help businesses build scalable, future-proof AI systems that drive real value and become integral to their operations.