AI Agents for Decision Support and Decision Augmentation

March 21, 2024

For external-facing customer services, many organizations already use AI automation. Customers of e:commerce sites received personalized recommendations “related to items you viewed” and instant credit checks. Customers of home camera systems received an automated alert of someone suspicious in their backyard. 

Humans oversee these results through quality control checks and customer feedback but are not “in the loop” of making every recommendation or alert. Gartner describes these use cases as “decision automation”, compared to “decision support” and “decision augmentation” where AI informs a human decision. 

Many people at work are already using GenAI to help write drafts of emails or strategy docs using natural language prompts. These same tools can provide numerical answers for questions such as “How many U.S. state capitals include the letter A in the city name?” 

There are fewer readily available options to chat with your organization’s data. In particular, what has been lacking in the data analytics marketplace is AI decision support and augmentation of data analytics for colleagues who are not data practitioners or data scientists. 

Microsoft has been working for several years on enabling natural language questions of data. Participants in the Microsoft Insiders program received beta access back in November 2019. Initially named the Ideas button, it is now labeled the “Analyze Data” button in Microsoft Excel. Analyze Data works best with clean, tabular data. 

Meanwhile, Microsoft Power BI offers a Q&A feature where analysts can ask questions of data in Excel, SQL Server, or Power BI datasets. It’s a nice shortcut to insert a chart or pivot table. A limited version, Power BI Free, is bundled in some Office 365 subscriptions. Power BI still requires knowledge and experience in how to use business intelligence (BI) software. 

Which brings us back to, excluding those of us with BI subscriptions and training, there has been a dearth of software that enables colleagues who are not data specialists to ask natural language questions of the company’s internal data alongside 3rd-party data enrichment. 

That’s where comes in. Asking questions in Patterns feels like communicating with a pleasant and informed human data analyst who works for your organization. That AI data analyst personalizes information delivery for what you want to know, when you want to know it. The AI learns from your interactions and feedback, continuously refining its recommendations to deliver the most pertinent information. 

Let's say you're a supply chain manager focused on optimizing inventory levels. By expressing your specific interests to Patterns AI, it can push insights about stock levels, demand forecasts, and potential bottlenecks directly to your email inbox or your Slack channel. This personalized approach not only increases the relevance and engagement of data insights but also saves valuable time and improves overall efficiency.

Or let’s say you work in finance and are looking into venture capital (VC) investments. You can ask the Patterns Vega bot questions like this one, and then continue to iterate, for example to drill down by specific U.S. state. 

Interactive AI Data Analysts

Have you ever stared at a blank screen, unsure of what questions to ask to extract value from your data? It's a common challenge in building AI interfaces, known as the "empty chat box" problem. If you flip this issue on its head, and instead prompt the AI to take the first action, the AI provides proactive suggestions and prompts based on your context and previous interactions, guiding you towards the insights you need.

The benefits of AI go beyond just generating initial questions. You can engage in natural language conversations with your AI analyst, asking follow-up questions and diving deeper into specific insights. The AI's contextual understanding allows it to provide meaningful responses and guide you through the data exploration process.

Interactive data visualization and guided analytics further enhance the user experience. AI can automatically generate visually compelling charts and graphs, highlighting key findings and making complex data easily digestible.

Integration with Data Catalogs, Semantic Layers, and Data Warehouses

To fully leverage the power of natural language AI in data analysis, it's helpful to ensure seamless integration with existing data infrastructure, such as data catalogs, semantic layers, and data warehouses. Data catalogs and semantic layers such as AtScale, DBT, and LookML are designed to provide a unified and business-friendly view of an organization's data, abstracting away the complexity of underlying data sources.

By integrating AI agents with semantic layers, organizations can enable natural language querying and contextual understanding across their entire data landscape. For example, a marketing analyst could ask, "What was the impact of our latest campaign on sales in the Northeast region?" The AI agent, powered by the semantic layer, would understand the business context and retrieve the relevant data from multiple sources to provide a comprehensive answer.

Data warehouses, such as Snowflake and Amazon Redshift, serve as centralized repositories for structured data, enabling efficient storage, processing, and analysis. The Patterns AI agent seamlessly integrates with these data warehouses, leveraging their scalability and performance to deliver real-time insights through natural language questions.

Moreover, the integration of AI with data catalogs, semantic layers, and data warehouses helps to enable a more unified and consistent view of data across the organization. This eliminates data silos and ensures that everyone is working with the same set of accurate and up-to-date information. With AI agents acting as the interface between people and the underlying data infrastructure, organizations can achieve a seamless and intuitive data analysis experience for all colleagues, not just data analysts and data scientists. This frees up time for data practitioners to focus on more specialized projects. 

Embracing the Future of AI Data Analysts

AI bots are enabling a new generation of data analytics. By harnessing the power of AI for data analytics, organizations like yours can unlock hidden insights, make data-driven decisions faster, and gain a competitive edge. 

Natural language analytics extend business intelligence to more parts of the organization with AI bots that feel like interacting with a coworker, while freeing time for data practitioners and data scientists to work on more specialized projects. 

As organizations embrace this new paradigm of AI-driven data analysis, they must also invest in the necessary infrastructure, talent, and processes to support it. This includes building robust data pipelines, ensuring data quality and governance, and fostering a culture of data-driven decision-making. Through this data journey, organizations can unlock the true potential of data and drive transformative business outcomes.

The future of data analysis is here, and it's powered by AI. Are you ready?

To learn more and discuss your organization’s specific use cases for AI natural language analytics, visit and pick a time to talk with us.

Analytics in natural language