November 2025 Summaries

Dialpad, an AI-powered communications intelligence platform, sought to overcome integration challenges by developing a system that improved configurability and scalability for its vast array of customer requests. Faced with the need to support 38+ integration requests and reduce the time taken to build native integrations, Dialpad embraced an AI-first architecture to enable rapid, agentic integrations. This shift was driven by their agentic AI initiative, which required a platform capable of dynamic orchestration by AI agents. After evaluating multiple vendors, Dialpad chose Membrane for its ability to handle multi-tenant architectures, provide MCP-native capabilities, and allow per-customer configurability. This choice facilitated the creation of more agentic integrations than traditional workflows, significantly reducing integration timelines from two months to two weeks. Membrane's design supports the vision of self-integrating applications, where AI agents can autonomously construct and maintain integrations in real-time, aligning with Dialpad's goal of a frictionless, on-demand integration system.

Software integration is undergoing a transformative shift with the advent of AI-driven systems that not only generate but also maintain production-grade integrations autonomously. This development promises to address a significant challenge faced by B2B companies: the labor-intensive and costly process of building and maintaining numerous software integrations, which often limits the number of supported applications. The introduction of tools like Membrane aims to revolutionize this landscape by enabling rapid generation of integrations with minimal maintenance, leveraging AI to handle complex tasks such as understanding API behaviors and managing infrastructure. This shift could fundamentally alter the economics of software connectivity, making it feasible to support a broad range of applications, including niche tools, and offering companies a competitive edge through extensive integration capabilities. While AI systems can handle many tasks autonomously, human oversight remains crucial for security and compliance, suggesting an evolving partnership between AI and human operators in managing software interoperability.

Membrane, formerly known as another entity, is focused on pioneering a future where software is capable of self-integrating, as highlighted in a rebranding announcement from their CEO and Founder, Daniil. The platform offers a comprehensive range of resources such as tips, case studies, guides, comparisons, and technical insights centered around integrations, providing valuable knowledge and unique perspectives from their team. The company encourages users to experience this innovative approach by booking a demo or trying the platform themselves.

Hireology, a company operating in the HR, hiring, and people platforms industry, faced challenges in scaling its integration capabilities due to a time-intensive, bespoke approach that hindered their ability to expand into the middle market. To address this, they adopted Membrane, a platform enabling the development of integrations that are both scalable and customizable, significantly reducing integration build times from 12 weeks to as little as six hours. Membrane's architecture allows Hireology to write integration logic once and apply it across multiple systems without further changes, using features like Universal Flows and a unified authentication framework. This shift has not only enhanced Hireology's operational efficiency but has also facilitated their upmarket expansion, allowing them to support a wide array of HR systems critical to their customers. The partnership with Membrane has been pivotal in Hireology's strategy, providing a foundation for future growth and the potential for further innovation with AI-driven integration capabilities that promise even faster and more autonomous integration development.

The Model Context Protocol (MCP) is an open protocol designed to enhance the integration of AI applications with SaaS tools by providing a standardized interface for agents to discover, access, and invoke external tools across various platforms. It operates on a client-server architecture, where MCP Servers expose capabilities as standardized tools, and MCP Clients—often embedded in AI applications—connect to these servers to utilize the available tools. This modularity allows for dynamic tool addition without altering the AI applications. MCP enables AI agents to orchestrate workflows across multiple SaaS applications, adapt to new integrations, and deliver a seamless, action-oriented user experience, addressing the modern expectation of effortless software interoperability. The Membrane MCP Server serves as a practical, open-source implementation of this protocol, offering features such as dynamic tool selection, session management, and secure access, thereby simplifying the process for developers and product teams to scale integrations without additional engineering efforts.

Membrane is an innovative AI-driven platform aimed at transforming the way product integrations are developed, by automating the process that traditionally requires extensive manual coding. The rebranding from Integration App to Membrane symbolizes a shift towards AI-powered integration generation, promising rapid deployment of production-ready integrations by interpreting natural language descriptions. Comprising three components—Membrane Agent, Membrane Engine, and Membrane Packages—it enables developers to create, manage, and deploy integration solutions efficiently and reliably, without being constrained by the need for deep expertise in various API technicalities. Membrane facilitates seamless scalability and adaptability to changes in APIs, while preserving full ownership and control over the integration code. By integrating into existing developer workflows and embracing cutting-edge AI capabilities, Membrane seeks to significantly reduce the time and resources spent on integration tasks, thereby allowing businesses to focus on enhancing their core products.

The article explores the differences between Unified APIs and embedded iPaaS, emphasizing the unique features and limitations of each. Unified APIs, such as those offered by services like Merge and Apideck, consolidate basic functions across different apps into a single API but struggle with complex, customized tasks. In contrast, embedded iPaaS integrates directly within applications, providing a more seamless user experience but often requiring individual setup for each integration, which can be cumbersome. Membrane is highlighted as a superior solution that combines the flexibility of iPaaS with the simplicity of Unified APIs, offering AI-powered integration tools that streamline the process and reduce the time and resources needed for app integration. Membrane's approach allows businesses to create, manage, and deploy integrations efficiently, with an emphasis on customization and global support, positioning it as a leader in the evolving landscape of SaaS integrations.

The SaaS industry is experiencing a transformative shift with the integration of AI agents, which are designed to automate workflows, interpret user intent from natural language, and connect disparate tools to streamline user experiences. One of the key challenges in this evolution is building these agents to work seamlessly across the vast landscape of SaaS integrations, traditionally limited to in-house development and maintenance. Membrane offers a solution by providing a universal integration layer that gives AI agents access to thousands of SaaS tools without the need for custom integration, enhancing user experience with dynamic tool selection, real-time updates, and reduced engineering overhead. This allows SaaS companies to deliver more flexible and powerful agent-driven features, lower maintenance costs, and ensure compatibility with emerging tools, ultimately driving competitive advantage in the market.

Anthropic's Model Context Protocol (MCP) is an open framework designed to streamline the integration of AI models with various business tools and systems, addressing the challenges of traditional custom integration methods. By providing a consistent and secure interface, MCP simplifies tasks such as accessing files, databases, and APIs, allowing AI models to leverage diverse resources efficiently. The Membrane MCP server, built on the same technologies, enhances this integration by enabling large language models (LLMs) to connect with thousands of tools and data sources in a unified manner, turning previously complex integration processes into straightforward workflows. This approach is likened to the transformative impact of TCP/IP in computer communications, aiming to create a universal language for AI model integration. The growing community around MCP, evidenced by the popularity of its GitHub repositories, signifies its expanding role in the future of digital integrations, with Membrane at the forefront of this movement.

In the next decade, the landscape of B2B technology is expected to shift dramatically as companies rely on thousands of specialized, seamlessly interconnected apps rather than the fragmented SaaS stacks seen today. The current challenges with integration complexity, where buyers demand seamless software interactions, highlight the necessity for a robust integration infrastructure to unlock the full potential of SaaS. Existing solutions like Embedded IPaaS and Unified APIs fall short due to scalability and flexibility issues. Membrane proposes a revolutionary integration layer leveraging AI to simplify and scale custom integrations, enabling developers to create numerous deep integrations efficiently. This approach is poised to redefine SaaS value by prioritizing seamless ecosystem compatibility, and it promises to drastically reduce the time required for deploying new applications. The advent of a universal integration layer marks a significant step towards an era where software self-integrates, enriching and enhancing the overall SaaS experience.

Sequence, a London-based platform that streamlines pricing, billing, and invoicing for B2B SaaS companies, replaced its existing unified API integration stack with Membrane, previously known as Integration App, to enhance its flexibility and customization capabilities. The switch was motivated by the need for real-time integrations and the limitations experienced with their previous provider, which included sync delays and inconsistencies across different applications. Membrane provided Sequence with a robust and consistent core across SaaS apps, real-time data syncing, and the ability to customize integrations per client or application. This transition allowed Sequence's engineering team to concentrate on core product development rather than constructing integrations from scratch, resulting in faster time-to-market for new features. With Membrane, Sequence successfully implemented over 20 different integration scenarios, including those with Xero, QuickBooks, Salesforce, and HubSpot, and plans to expand further into sales, document management, and e-signature tools.

As the integration landscape evolves with the rise of AI, traditional methods of building integrations—relying on iPaaS tools, custom-coded connectors, and manual processes—are becoming insufficient to meet the demands of modern software ecosystems. Customers now expect seamless connectivity across a wide array of tools, necessitating a shift from rigid, single-use connectors to scalable, AI-driven integration systems. AI is transforming how integrations are developed by enabling dynamic adaptation to different schemas, debugging issues, and generating app-specific variants of core use cases. This requires a new architecture that separates business logic from app-specific implementations, supports runtime execution, and facilitates AI participation in the integration lifecycle. Membrane offers a universal layer that allows teams to manage and scale integrations efficiently, using AI to automate and optimize the integration process, thus reducing complexity and manual overhead. This approach not only enhances the flexibility and scalability of integration systems but also prepares organizations for future advancements where AI agents autonomously manage and adapt integrations.

Storyblok, a headless content management system (CMS) platform founded in 2017, has successfully leveraged Membrane to enhance its integration infrastructure, significantly reducing the time required to develop and deploy new integrations from two weeks to just one day. This strategic move was driven by Storyblok's aim to improve connectivity with various SaaS applications, a necessity given their ambition to integrate with numerous new apps annually. Membrane was chosen over other providers like Tray.io and Nango due to its robust features, such as a flexible visual editor, abstraction layers for authentication, and configurability, allowing Storyblok engineers to focus on building business logic rather than backend infrastructure. Notably, Membrane enabled Storyblok to quickly build a custom Algolia connector and manage complex integrations like Shopify, which required unique authentication mechanisms. This partnership has not only increased productivity and market speed for Storyblok but also positioned Membrane as a leading integration infrastructure provider for other established companies.

Sembly AI, a smart meeting assistant founded in 2019 and headquartered in New York, has successfully scaled to over 40 integrations using Membrane's modern infrastructure, bypassing the limitations of traditional embedded iPaaS tools. This has enabled Sembly to implement a single OAuth integration for all apps, facilitating seamless and efficient connectivity across platforms such as CRMs, productivity tools, and documentation tools. By leveraging Membrane's capabilities, Sembly has enhanced its go-to-market strategy, gaining credibility and expanding visibility through co-marketing opportunities and marketplace listings, while also preparing for future bi-directional integrations to enrich AI experiences. The integration strategy has positioned Sembly as a competitive player in the AI and productivity SaaS industry, allowing it to focus on business logic and scalability without extensive custom endpoint coding.