June 2025 Summaries

Modern data architecture represents a shift from traditional data systems, which are increasingly unable to handle the vast amounts of data generated daily. To address this, organizations are adopting cloud-native architectures that offer speed, scalability, and improved data quality, essential for informed decision-making. Key components of modern data architecture include batch and real-time data ingestion, utilizing data lakes and lakehouses for storage, and implementing robust data catalogs for easy data discovery and management. Data transformation and analytics are emphasized, with tools like dbt and Apache Spark facilitating the process, while security and governance remain central to protecting and organizing data. Additionally, the implementation of real-time event pipelines, federated governance, and cloud-native approaches further enhance modern data systems. Emphasizing advanced analytics, modern architectures integrate seamlessly with analytic and business intelligence tools, promoting data usability and business growth. The move towards modernization is not merely about replacing legacy systems but involves comprehensive changes to achieve better scalability, efficiency, and alignment with business goals, often through phased implementation strategies.

In an era where traditional product analytics tools fall short, modern product analytics emphasize quick, actionable insights into user behavior to enhance product performance. This involves tracking user actions like signups and purchases to reveal patterns that indicate feature adoption and churn signals. Unlike broader data analytics or static business intelligence, product analytics focus on in-app behavior to optimize product experiences, informing product teams' decisions. Tools like Hex facilitate this by offering no-code solutions for event tracking, cohort analysis, funnel analysis, and predictive modeling, which help identify conversion barriers and retention opportunities. By consolidating data from various sources and automating analytics pipelines, teams can create interactive dashboards and reports that provide real-time insights, empowering all stakeholders to make informed decisions. Hex also addresses data silos by integrating analytics within product development, allowing for collaborative exploration of user experience data, ultimately driving continuous growth and innovation.

In 2025, data teams are navigating a rapidly changing landscape driven by the evolution of AI, tool acquisitions, and heightened stakeholder expectations, as revealed in a comprehensive survey involving over 2,000 data professionals. Despite the excitement surrounding AI, only a small fraction of data teams prioritize its implementation due to concerns about readiness and trustworthiness, prompting a call to action for embracing AI to stay competitive. Additionally, frustrations with self-service BI solutions are prompting 70% of teams to reconsider their technology stacks, as traditional dashboard approaches fail to address the ever-evolving questions and challenges posed by stakeholders. Despite fears of AI replacing jobs, data teams are actually expanding, with 90% of respondents maintaining or increasing team size, highlighting the ongoing need for human judgment and collaboration alongside new technological tools. This survey underscores the challenges and opportunities facing data teams, emphasizing the importance of adapting to technological advancements while leveraging human expertise to drive strategic value within organizations.

Christian Kleinerman, Executive Vice President of Product at Snowflake, discusses the evolving landscape of data and AI during a fireside chat at the Snowflake Summit. Reflecting on his career from SQL Server development at Microsoft to managing infrastructure at YouTube, Kleinerman emphasizes the importance of using a product to understand its strengths and weaknesses fully. He acknowledges underestimating AI's impact, noting that the technology is more advanced than its current applications, and highlights the potential of AI in enterprise data management, especially concerning security and governance. Kleinerman also discusses Snowflake's strategic emphasis on data portability, openness with Iceberg, and the significance of partnerships to adapt to market needs. With the rapid advancement of AI, he remains cautious about making future predictions but sees significant opportunities in aligning AI developments with enterprise data needs. Additionally, he addresses customer concerns about AI model capacities and the impact of AI on employment, stressing the importance of balancing technological advancements with social considerations. Snowflake's cost management strategies and efforts to simplify infrastructure tasks are also discussed, alongside advice for startups to gain traction by leveraging clear use cases and aligning with Snowflake's sales strategies.

Scaling secure and personalized client insights is crucial for Candidly, a mission-driven company helping modern workforce accelerate and optimize student debt repayment. To achieve this, the company uses Hex embedded analytics to provide in-depth user-friendly perspectives into the impact their platform generates. Founded in 2016, Candidly has generated nearly $2 billion in projected student debt impact and helped users reduce their student loan repayment time by a projected 230,000 years. By leveraging Hex's seamless and secure embedding capabilities, row-level security, and good data organization, Candidly can deliver accessible and engaging insights to its clients and partners, fostering stronger collaboration between internal teams and empowering faster, more informed actions.

The Snowflake Summit featured a packed week with fresh perspectives on the role of agents, semantic layers, and data governance. The event showcased successful strategies from teams at the NBA, HubSpot, Ramp, and Chime, including focusing on organizational influence, collaboration, and speed to build relationships. Key takeaways include recognizing that measuring data team ROI is a myth, prioritizing people over tools, and building with the goal of publishing and sharing work. The summit also highlighted the importance of distribution, politics in data jobs, and focus on new, novel, and gnarly challenges. Additionally, partnerships between Hex and Snowflake were announced, including direct support for Snowflake's Semantic Views inside of Hex.

In an effort to surround himself with valuable advice and guidance, a CEO attended a dinner event where he met Linda Tong, who was then the CEO of Webflow. He was impressed by her experience leading product at the NFL, serving as a GM at AppDynamics, and now as CEO of Webflow. The CEO asked Linda to join his company's board, citing her unique blend of product instinct, go-to-market savvy, and real-world leadership skills. Linda brings a rare combination of skills that make her an ideal fit for helping the company navigate its next era, including her understanding of what it takes to build great products and enduring companies.

Modern data platform architecture requires flexibility and scalability to efficiently manage the influx of new data sources, integration of various tools, and evolving business requirements. The architecture should consist of layered systems that include ingestion, processing, transformation, analytics, consumption, governance, and security, each serving a specific function to maintain data integrity and usability. The ELT approach is commonly used, with data loaded in raw form before transformation, offering flexibility and ease of debugging. Governance and security are crucial for protecting data assets, and strong governance practices enhance performance and trust. Different storage solutions like data warehouses, data lakes, and data lakehouses cater to various data needs, and choosing the right one depends on data format and use cases. As complexity grows, data mesh and data fabric architectures help manage it by decentralizing data ownership and creating a unified layer for data discovery and integration. The choice between batch and streaming data pipelines depends on latency needs, and a combination of cloud and hybrid deployments can balance performance, security, and compliance.

Snowflake has introduced Semantic Views, a new way to define business metrics and logic directly in your database. This integration allows users to explore and analyze trusted measures, dimensions, and joins defined in Snowflake from Hex's no-code Explore experience. With this integration, data teams can bring trusted context into Hex for a single source of truth, making it easier to deliver consistent insights without compromising on governance or accuracy. The new integration also enables the use of Ask Magic in Explore, a conversational AI interface that leverages governed metrics and business logic from synced semantic models to generate more accurate outputs. This makes it possible for users outside of the data team to access reliable answers, democratizing data and speeding up analytical workflows.

Snowflake Summit is an annual conference that brings together thousands of people from various industries, including AI and tech. The event's theme this year is "Build the Future of AI and Apps," which aims to help organizations rethink problems they thought they had solved 20 years ago with the help of AI. While some attendees are genuinely interested in learning about the latest advancements in AI, others are attending for parties and free drinks. To separate the signal from the noise, it's essential to focus on real value rather than just hype and trends. Key topics at this year's summit include Apache Iceberg and open formats, which enable modern teams to do more with less lock-in; compliance and access controls in AI-embedded workflows; and smart governance that enables both AI and security. The event will also feature various sessions and speakers from well-known companies like Hex, OpenAI, and the NBA, showcasing their approaches to integrating AI into end-to-end workflows and building data stacks that stand the test of time.