Frequently Asked Questions

Conversational analytics allows people to work with data using plain language rather than code, dashboards, or predefined filters. Instead of navigating charts or writing SQL, a user asks a question in natural language and the system determines how to answer it.

Behind the scenes, the platform interprets intent, identifies relevant data, performs the analysis, and returns results in a structured form such as tables, charts, and written explanations. The experience is iterative, meaning users can ask follow up questions to refine or expand the analysis without starting over.

‍

Agentic workflows is a series of agents, each responsible for a step in a workflow process. In Lumi, this includes using an orchestration of agents for interpreting intent, gathering context, writing query logic, validating results, and formatting outputs.

A semantic layer is a business friendly abstraction that sits on top of raw data and translates technical data structures into consistent, reusable business concepts. In Lumi, this is one of the roles of the Knowledge Base. It stores field definitions, table relationships, naming conventions, and business context. This allows users to ask questions in business terms without needing to know how data is structured underneath.

‍

No. Lumi is useful for repetitive ad hoc query requests, but it still depends on data teams to model data correctly, define metrics, and maintain the Knowledge Base. By handling routine analytical questions, Lumi reduces data bottlenecks and allows data teams to focus on enablement, governance, and higher impact work rather than constant query fulfillment.

No. Dashboards are useful for recurring reporting. Lumi complements them by answering ad hoc questions and performing deeper analysis without the need for technical SQL or Python skills, unlocking operational insights that would otherwise be buried.

Lumi is a chat-based agentic data analyst that uncovers hidden ROI across ERPs and databases faster and at a fraction of the cost of manual analysis.

• Chat: The Chat module is the primary interface for interacting with data using natural language. Users type questions in normal business language, and Lumi translates those questions into executable SQL, runs the queries, and returns structured results as tables, charts, or written summaries. The Chat interface also shows the underlying SQL code and explanations of the steps taken to generate results.
• Boards:  Boards are persistent spaces where users can pin insights, charts, and analyses for ongoing tracking and sharing. Unlike one-off responses in Chat, Boards allow teams to curate and organize key reports and dashboards, schedule refreshes, and keep insights up-to-date without re-asking the same questions manually.
• Knowledge Base: The Knowledge Base is the semantic and governance module behind Lumi. It defines how Lumi understands the business data by storing connection details, field descriptions, table schemas, relationships, metric definitions, business context, and aliases. This semantic layer ensures that natural language questions are interpreted accurately and consistently according to the organization’s rules and definitions. Other modules like Chat and Boards depend on a configured Knowledge Base to generate correct results.
• Human Verification: Human Verification is the module that enables human in the loop oversight for insights generated by Lumi. It allows business users to request verification on insights or analyses produced in Chat or saved to Boards. Designated reviewers can examine the underlying logic, data sources, and calculations, then confirm, reject, or comment back on the insight with feedback or corrections. This creates a structured review workflow ensuring results are trusted, validated, and aligned with business definitions before being shared or operationalized. Human Verification allows organizations to scale access to insights while maintaining accountability and governance.

Lumi is built for enterprise environments with diverse data users, including:

• Business Users and Decision Makers: People who need quick answers from data without relying on SQL or BI tool expertise.
• Data Analysts and BI Teams: Teams that want to reduce repetitive query requests, accelerate prototyping, and validate results quickly with transparent SQL and logic.
• Operational Teams (Supply Chain, Finance, Sales, Marketing): Functional teams that benefit from instant insights into performance variances, trends, and root causes without waiting for scheduled reports.
• Executives and Leaders: Leaders who need on-demand insights to make strategic decisions without bottlenecks in analytics workflows.

Across these roles, Lumi helps reduce friction between asking questions and getting actionable answers by lowering technical barriers and automating routine analysis.

By eliminating manual analytic processes, Lumi allows organizations to:

• Instantly Uncover Hidden Value. Find insights that drive financial results within seconds.
• Boost Productivity. Make better, faster data-driven decisions.
• Reduce Time to Insights. Automate data analysis and insight extraction.
• Free up Technical Teams. Redirect focus from ad-hoc requests to strategic initiatives.
• Save on Hiring Costs. Produce more analytics without needing to hire.

‍

Lumi connects to a wide range of ERPs, databases and data warehouses, including SAP, Postgres, and Snowflake. You can access the full list of connectors here.

Yes. Teams can connect spreadsheets alongside their other data sources. This allows Lumi to generate insights that combine structured and unstructured data in a single analysis.

No. Lumi can connect directly to operational systems such as SAP, Oracle, and Microsoft Dynamics.

Lumi shows the logic, assumptions, and calculations behind every result. Users can inspect queries, review intermediate steps, and trace how conclusions were reached.

For decisions where additional assurance is required, Lumi has a built in module called Human Verification. This allows teams to request a manual review of an analysis. During verification, logic and assumptions are reviewed to confirm they follow approved business definitions and analytical standards. Results can be approved, rejected, or sent back with feedback, creating a clear audit trail.

This combination of transparent logic and optional human verification ensures insights are not just fast, but trusted, repeatable, and suitable for high-impact operational decisions.

‍

Lumi asks clarifying questions or breaks the request into smaller steps before running analysis. This reduces incorrect assumptions and improves accuracy.

Lumi runs inside your environment and does not move or store your data externally. All processing happens within your network and follows your existing security controls. Lumi supports role-based access control, enforces query cost and duration limits, and connects through advanced data gateway connectors designed for secure, read-only access. Lumi is also SOC 2 compliant so your data remains protected end to end. Learn more about our security page.

No. All queries are processed within the customer’s environment. No data is copied or duplicated.

Yes. Admins manage sign-in permissions (SSO), data access, and visibility access across users and Knowledge Bases. Lumi is purpose built with expected enterprise grade data security and controls.

Yes. Admins can set limits on query cost and query duration, depending on the source system.

Most tools focus on query generation. Lumi takes on the broader workflow a competent data analyst is expected to manage. It asks clarifying questions when inputs are vague, checks for errors before a query runs, and summarizes the findings in plain language. It explains the assumptions behind each step, shows the logic it followed, produces the visuals needed for reporting, and points to practical next actions. This mirrors the full workflow of a competent analyst instead of stopping at code generation.

Microsoft Copilot operates inside Power BI and is designed to help users work faster within existing reports, visuals, and semantic models. It assists with generating charts, summaries, and explanations based on data that has already been modeled and published in Power BI.

Snowflake Copilot is primarily developer focused features within the Snowflake ecosystem. Its  designed to help developers generate SQL, or code based on text input.

Traditional BI tools excel at tracking predefined metrics through dashboards. They’re not designed to answer the highly specific, ad-hoc, or exploratory questions teams face daily. Lumi lets business users ask custom, plain-language questions and instantly pulls the right data, runs the analysis, and explains the logic behind the answer. Lumi works alongside tools like Tableau and Power BI by handling the long tail of exploratory and one-off questions that dashboards aren’t built to serve.

No. Text-to-SQL is only one small piece of what Lumi does. Lumi uses multiple agents, a structured knowledge base to interpret intent, apply business definitions, run the right analyses, and generate clear, decision-ready insights. SQL generation may happen behind the scenes, but the value comes from understanding context, automating analysis, and explaining results, not simply converting text into queries.

Lumi answers a wide range of simple, complex, and even vague questions. Clients use it to surface exceptions, diagnose root causes, and uncover opportunities hidden in large structured datasets. Click here to learn more.

Mid-to-large CPG and F10 retail clients have identified multi-million dollar opportunities across inventory, sales, and procurement. Click here to view our case studies.

‍

Onboarding is swift, typically requiring just a few hours of effort. The process involves (1) establishing a connection and (2) configuring the knowledge base. We also offer personalized, white-glove onboarding for enterprise customers.

Yes, we offer a low-risk, low-effort pilot program that allows enterprise customers to trial Lumi's capabilities. It’s designed to deliver value within ~2 weeks through a focused scope and defined use cases. Learn more here.

Billing questions can be sent to billing@lumi-ai.com.