2024 Semantic Layer Innovations for Enterprise Analytics and Generative AI

2024 Semantic Layer Innovations for Enterprise Analytics and Generative AI

As we step into the realm of 2024, the data and AI landscape is poised for further transformation, driven by technological advancements, market trends, and evolving enterprise needs. In this blog post, three AtScale executives—Dave Mariani, CTO and Founder; Elif Tutuk, VP of Product; and John Langton, VP of Engineering—share their reflections on 2023 and unveil their predictions for the coming year.

Dave Mariani: Navigating the Landscape of Generative AI and Cloud Migration

2023 in Review:

This past year marked a period of cautious exploration of generative AI within the enterprise landscape. While the promise of this technology remained enticing, practical IT investments were relatively modest. Enterprises grappled with concerns over data residency, security, and the complexities involved in operationalizing cutting-edge technologies. As a result, interest in generative AI largely manifested in applications aimed at enhancing customer service through sophisticated chatbots and providing support to professionals in various industries.

Large Language Models (LLMs), a critical component of generative AI, continued to be under the control of major cloud vendors, with OpenAI leading the market. Google’s Bard and Amazon’s Anthropic made strides but struggled to catch up to OpenAI’s first-mover advantage. In-house development of LLMs by individual enterprises remained rare due to the substantial expertise, resources, and data required for training and maintenance.

Cloud migration was a dominant trend in 2023, but organizations approached it with a more cost-conscious lens. Databricks and Snowflake continued their dominance in the cloud analytics arena, with Databricks’ lakehouse architecture gaining traction. Google BigQuery maintained its stronghold within Google Cloud Platform (GCP) deployments. However, the economic headwinds led organizations to scrutinize the total cost of ownership more closely, favoring the more cost-effective lakehouse design pattern over the traditional data warehouse architecture.

2024 Looking Forward:

As we step into 2024, the cautious exploration of generative AI is expected to persist. While interest remains high, practical IT investments will likely remain modest, driven by concerns over data residency, security, and the operational complexities of implementing cutting-edge technology. Generative AI applications are anticipated to find their place in enhancing customer service and supporting professionals in various industries, serving as copilots in decision-making processes.

The dominance of LLMs is expected to stay centralized among a few key players, with OpenAI positioned to lead the space throughout 2024. Cloud migration will continue, with organizations adopting a more prudent fiscal outlook. The lakehouse design pattern is poised to gain further momentum, aligning well with the trend toward decentralized and flexible data management approaches. Microsoft Fabric is expected to garner attention as a contender in the data lakehouse movement, leveraging DirectLake technology to revamp data analysis processes.

Elif Tutuk: FinOps, Semantic Layers, and the Human Element

2023 in Review:

In 2023, the Financial Operations (FinOps) trends highlighted a heightened focus on balancing data delivery costs with the value generated. Organizations prioritized cost optimization strategies, streamlining expenses associated with data storage, processing, and analysis. AtScale’s innovation of an analytics usage dashboard, powered by the active metadata of the universal semantic layer, emerged as a game-changer. This tool provided organizations with visibility into analytics usage, helping them strike a balance between value and cost in cloud data warehouse analytics.

Semantic layers and metrics stores gained prominence in addressing the challenge of consistency across metric definitions. These solutions allowed users to create and define business metrics, governing them from data warehouses to serve downstream analytics, data science, and business applications. Additionally, composable data and analytics became essential as organizations sought flexibility in assembling and reassembling data and analytics capabilities for faster insights and proactive decision-making.

2024 Looking Forward:

As we look ahead to 2024, there will be a big emphasis on keeping the human element in the loop, particularly as generative AI gains momentum. Decision Intelligence becomes crucial, driven by the acceleration of research and adoption of composite AI models.

Two forces are expected to shape a new market around decision intelligence platforms: the combination of AI techniques such as natural language processing, semantic layers, and machine learning, and the convergence of technology clusters around composite AI, smart business processes, insight engines, decision management, and advanced personalization platforms.

Responsible AI takes center stage in 2024, addressing ethical choices in AI adoption, especially with GenAI. The emphasis will be on ensuring business value while mitigating risks, encompassing aspects such as transparency, fairness, bias mitigation, explainability, sustainability, accountability, safety, privacy, and regulatory compliance. We should expect to see the rise of new roles, like AI stewards, responsible for ensuring that AI technologies are implemented and used in a manner that aligns with ethical standards, legal compliance, and the organization’s values.

Furthermore, the data as a product trend continues to evolve, with data and analytics product management turning data into reusable products tailored to specific business outcomes.

John Langton: Technical Advancements and Future Integrations

2023 in Review:

In the technical realm of 2023, AtScale witnessed significant advancements. The integration of containers marked a milestone, enabling deployment on laptops, any cloud marketplace, or organizations’ own servers. The combined code and GUI editing capabilities were introduced as a first-in-market feature, providing users with versatile options for modeling.

Semantic Modeling Language (SML) emerged as a new standard for semantic modeling, not tied to a specific tool.

Complete integration with git version control, coupled with CI/CD enablement for data modeling, enhanced AtScale’s capabilities. More authentication options, expanded integration with tools and Identity Providers (IDPs), data catalog integrations, and improved telemetry for attribution and governance were instrumental in strengthening AtScale’s position in the technical landscape.

2024 Looking Forward:

As we step into 2024, AtScale is introducing new features and enhancements to fortify our platform’s reliability to continue to ensure uninterrupted access to critical data and analytics, even in the face of unforeseen challenges.

But the journey into 2024 is not solely about strengthening our foundations; it’s also about pushing the boundaries of innovation. AtScale is poised to introduce a suite of transformative features that will reshape the landscape of data and analytics:

  • Integration with Generative AI and Large Language Models (LLMs): Embracing the cutting-edge realm of advanced machine learning, this integration will seamlessly incorporate generative AI and LLM capabilities into the AtScale platform, unlocking new possibilities for data exploration and analysis.
  • MDX Generation: AtScale is introducing a new approach to query language flexibility with MDX generation by using LLM. This enhancement will empower users to interact with their data in more nuanced and human friendly ways, fostering a richer analytics experience.
  • Natural Language Interfaces: In a bid to democratize data access, AtScale is bringing natural language interfaces to the forefront. Users will be able to interact with the platform using natural language queries, breaking down barriers and making data and analytics accessible to a broader audience.
  • Built-in AI Capabilities: AtScale is taking a bold step towards comprehensive data set analysis by embedding built-in AI capabilities. This includes advanced functionalities such as forecasting measures, trend tracking, and anomaly highlighting, directly enhancing the decision-making process for users.

2024 promises to be a year of continued innovation and evolution for data, analytics and AI – both for enterprise and at AtScale. Get ready for 2024 and watch Dave and Elif discuss The Future of the Universal Semantic Layer in the Modern Data Stack.

GigaOm Sonar Chart - semantic layers and metrics stores