June 2024 Summaries

Retool has launched "Retool for External Apps," a new solution designed for building and managing apps for users outside an organization, now available on Business and Enterprise plans. This platform extension allows businesses to create customizable, secure, full-stack portals for external users such as customers, vendors, and partners, with features like built-in login, authentication, access controls, and audit logs. Users can whitelabel the interface and customize user journeys, including sign-in and account management processes, while leveraging Retool's granular, role-based permissions model to ensure data security. The platform also supports embedding Retool apps in existing dashboards, SaaS tools, or web apps, and offers user-based pricing tiers that scale with the number of external users, providing cost predictability as businesses grow.

A technical account manager at Retool provides insights into implementing cascading filters within the Retool platform to enhance data exploration tools, particularly for engineering teams. This process involves using Retool's intuitive interface and JavaScript capabilities to create a dynamic filtering experience by narrowing down data through hierarchical selection criteria. For instance, selecting a country updates state and city options accordingly, facilitating efficient navigation of large datasets. The guide outlines how to set up cascading filters using components such as Select dropdowns and Tables, with a focus on JavaScript logic to update filter options based on user selections. This approach is applicable to various hierarchical datasets, enabling teams like HR or support to refine data views efficiently. The document also highlights how the completed application can be downloaded and imported into a Retool instance, encouraging users to explore and implement cascading filters in their projects.

In 2024, AI technology continues to be seen both as a tool with significant potential and a subject of skepticism among tech professionals, with many viewing it as slightly overrated despite its rapid proliferation. A survey of around 750 individuals across various industries and roles revealed that while adoption rates are moderate, with sectors like consulting, real estate, and consumer goods leading in AI implementation, many companies still face challenges in achieving transformative impacts. The majority of respondents use AI tools such as copilots weekly, reporting productivity gains, especially with daily use. However, AI's integration into workplaces is not without issues, as concerns about data security, resource allocation, and model reliability persist. Furthermore, while leadership is generally keen on increasing AI investments, there remains a disparity between enthusiasm and practical application, with some individuals still using AI covertly due to unclear company policies. The study also highlighted that while AI-powered chatbots are prevalent, they may not represent the most innovative use cases, as many organizations are exploring internal applications like automating workflows and writing code. Despite the mixed sentiment, there is optimism about AI's future utility and a belief that its role in both internal and external business functions will expand as the technology matures.

At a recent event in London, hosted by Retool and OpenAI, members of the European tech community gathered to discuss the development and implementation of custom AI-powered applications. The event highlighted the increasing interest from developers and businesses in using AI to enhance operations, with many companies beginning with automation through chatbots and AI-powered internal tools. Dom Grillo from OpenAI noted that while the potential of AI is widely recognized, concrete strategies are often lacking, leading companies to develop custom solutions on top of the OpenAI API. Dobo Radichov from Holland and Barrett showcased an application that uses large language models (LLMs) to efficiently route users to data-driven answers. Retool demonstrated AI applications that improve customer support by using AI to manage simple queries and translate natural language questions into SQL queries for complex data warehouse navigation. The consensus was that custom AI applications offer more flexibility and relevance to specific business needs compared to off-the-shelf solutions, emphasizing the importance of adaptability in the rapidly evolving tech landscape.

Retool's latest Developer Day highlighted several new features and enhancements aimed at improving user experience and application functionality, alongside a preview of upcoming releases. In Q2, updates included the introduction of Retool AI capabilities, enhanced UI, enterprise and collaboration tools, and testing improvements. Notable features include dev mode for custom components, enabling instant hot reloading of changes, and the introduction of multipage applications for improved navigation and load times. The platform also supports custom product pages, integration with Amazon Bedrock for AI model connectivity, and internationalization for cross-regional app adaptability. Additional enhancements involve webdriver testing for application reliability, real-time audit log streaming to Datadog, and telemetry collection for self-hosted environments. Retool also introduced split pane layouts for more flexible design, editable key-value components, collaborative branches for team work, and form-building capabilities. Upcoming features teased include dynamic sidebars, theme customization, a revamped grid layout with Stacks, and error monitoring for real-time updates. Users are encouraged to upgrade to the latest release to access these advancements.

Enterprise AppGen introduces AI-powered app generation that promises to be fast, secure, and production-ready, but deploying AI models like pre-trained large language models (LLMs) requires additional refinement to meet specific business contexts. The text explores several techniques to enhance AI models, such as Retrieval-Augmented Generation (RAG) which combines retrieval models and generation techniques to provide context-specific responses, and fine-tuning, which adapts a model to specific domains by training it on a domain-specific dataset. Prompt engineering is discussed as a crucial skill for crafting effective AI prompts to generate desired outputs. Model switching is mentioned as a method to find the best AI model for a given use case by testing and comparing different models. Multimodality, the ability to generate both text and images, is highlighted as a way to create engaging multimedia content. Retool is presented as a platform that facilitates these enhancements, allowing for easy integration, manipulation, and testing of AI models to improve their performance and applicability.