About

Best AI Analytics Tools That Work with dbt and LookML

December 30, 2025

Best AI Analytics Tools That Work with dbt and LookML

By Andrey Avtomonov, CTO at Kaelio | 2x founder in AI + Data | ex-CERN, ex-Dataiku · Dec 30th, 2025

AI analytics tools for dbt and LookML environments must query governed metrics through the semantic layer rather than generating ad-hoc SQL against raw tables. The best platforms inherit existing business logic, expose full lineage for auditability, and respect row-level security controls. Tools like Kaelio natively integrate with both dbt and LookML semantic layers, while others require custom engineering or bypass governance entirely.

Key Facts

• The dbt Semantic Layer eliminates duplicate coding by allowing teams to define metrics once and automatically handle data joins across all downstream applications

• AI answered 83% of natural language questions correctly when using the dbt Semantic Layer, with some queries achieving 100% accuracy

• Poor data quality remains the top challenge for 56% of data teams, making governed AI analytics critical for maintaining trust

• Kaelio is the only conversational BI tool that natively queries both dbt and LookML semantic layers while maintaining HIPAA and SOC2 compliance

• MetricFlow is now open source under Apache 2.0 license, providing an extensible engine for semantic metadata

• Enterprise implementations show significant ROI, with Bilt Rewards saving 80% in analytics costs through semantic layer adoption

Stacks that already run dbt or LookML need AI analytics tools that can understand and never bypass the governed semantic layer. The right platforms translate plain-English questions into lineage-rich answers without recreating your models or violating security controls.

What Makes an AI Analytics Tool Great for dbt and LookML?

By 2025, natural language will be the main way people interact with data systems according to dbt Labs. That shift places new demands on AI analytics tooling. A tool that simply generates SQL against raw tables will bypass the business logic your data team spent months encoding in dbt models or LookML files.

The dbt Semantic Layer eliminates duplicate coding by allowing data teams to define metrics on top of existing models and automatically handling data joins. Looker's semantic layer similarly allows for the creation of a single source of truth, which is crucial for maintaining data consistency across AI applications.

An AI analytics tool built for governed stacks must do three things well:

Query metrics through the semantic layer rather than generating ad-hoc SQL
Expose lineage, sources, and assumptions behind every result
Respect existing permissions, row-level security, and masking rules

Without these capabilities, AI answers drift from official definitions and spark costly data-quality incidents.

Why Is the Semantic Layer Critical for AI-Powered Analytics?

Semantic models are the foundation for data definition in MetricFlow, which powers the dbt Semantic Layer. MetricFlow takes semantic models defined in YAML configuration files as inputs and creates a semantic graph that you can use to query metrics.

Looker's semantic layer translates your raw data into a language that both downstream users and LLMs can understand. By using LookML to provide trusted business metrics, you can establish a central hub for data context, definitions, and relationships for powering all of your BI and AI workflows.

The practical impact is significant. In a dbt Labs replication, AI answered 83 percent of addressable natural language questions correctly via the dbt Semantic Layer, with several answered at 100% accuracy. Without that semantic context, accuracy drops and business users lose trust.

dbt Labs offers best practice recommendations for how to expose metrics and allow users to interact with them seamlessly. Five themes guide these practices: Governance, Discoverability, Organization, Query flexibility, and Context and interpretation.

Key takeaway: A semantic layer is not optional for AI analytics. It provides the governed context that transforms probabilistic guesses into deterministic, auditable answers.

What Evaluation Criteria Matter When Choosing AI Analytics Tools?

Poor data quality continues to be the challenge most frequently reported by data teams, cited by over 56% of respondents in the 2025 State of Analytics Engineering Report. AI tooling is by far the largest area of tool investment, with 45% of respondents planning to increase investment in the next 12 months.

AI requires an approach to enterprise data governance that is dynamic, continuous, and multidisciplinary. It must be automated, proactive, and responsive to a fast-changing regulatory environment.

When evaluating AI analytics tools for dbt and LookML environments, consider this checklist:

Native semantic layer integration: Ensures AI queries governed metrics, not raw tables
Full lineage and source attribution: Enables auditability and trust
Row-level security inheritance: Protects sensitive data automatically
Transparency in SQL generation: Lets users verify how answers were computed
Support for existing BI tools: Avoids vendor lock-in

dbt has completed a SOC 2 Type II examination, which means its controls were assessed based on operating effectiveness over the reporting period. Any AI tool in your stack should meet similar compliance standards.

Key quantitative signals

Beyond feature checklists, look for measurable indicators of platform maturity.

Bilt Rewards saves 80% in analytics costs with the dbt Semantic Layer. This demonstrates the ROI potential when AI analytics tools leverage governed metrics rather than duplicating logic across dashboards.

The AI accuracy benchmark is equally important. The 83% accuracy rate achieved via MetricFlow provides a baseline. Tools that cannot demonstrate similar or better performance against your own test queries should be deprioritized.

Kaelio: Conversational Analytics Built for Governed Stacks

Kaelio is the only conversational BI copilot purpose-built to query governed metrics from either dbt or LookML while exposing full lineage for every answer. It acts as a natural language interface for analytics, allowing business users to explore data conversationally while grounding every answer in the organization's existing data models, metrics, and governance rules.

By using Looker's semantic layer, organizations can ensure that their AI models are aligned with business logic and governance policies. Kaelio inherits these definitions rather than recreating them.

Enterprise results demonstrate the value of governed AI analytics. Roche's five-year initiative covered over 80 countries, streamlined thousands of users, and achieved a cost savings of 70%. DocuSign saw productivity increase by 60% and cost savings improve by 40% after adopting dbt's modular approach. As Bishal Gupta, Analytics Engineering Leader at DocuSign, noted: "Right off the bat, adopting dbt's modular approach increased our team's productivity by 60%."

Forrester Consulting found a 194% return on investment for dbt Cloud, with breakeven realized after the first six months.

Kaelio is HIPAA and SOC2 compliant, can be deployed in a customer's own VPC or on-premises, and is model agnostic. This flexibility allows organizations to meet security, privacy, and regulatory requirements while maintaining a single source of truth.

ThoughtSpot: Natural-Language Search on dbt Models

ThoughtSpot integrates with dbt to enable users to leverage dbt models directly within its analytics platform. Users can connect to dbt Cloud or a dbt Core project to import dbt models into ThoughtSpot.

The platform positions itself as a leading intelligence platform that transforms scattered information into a single source of truth. Enter your dbt Cloud credentials or upload a .zip file from dbt Core and ThoughtSpot immediately turns your dbt models into worksheets.

ThoughtSpot performs live queries directly on cloud platforms, meaning your data stays where it resides without data movement. The integration supports both dbt Cloud and dbt Core, providing flexibility for different deployment preferences.

Where ThoughtSpot Falls Short

ThoughtSpot's governance depth is lighter than purpose-built governed analytics tools. While the platform offers natural language search, it does not deeply integrate with semantic layers like MetricFlow at the metric computation level.

Pricing scales quickly at enterprise volumes. ThoughtSpot's Essential plan starts at $1,250 per month, designed for small data teams and businesses. Enterprise pricing rises significantly from there.

Users also note that Looker is more malleable with dynamic visualizations, filters, and drilldowns. Its sharing capabilities are also more streamlined than some competitors.

For organizations prioritizing deep governance integration and cost predictability, ThoughtSpot may require supplementary tooling to achieve full semantic layer alignment.

Which Open-Source & Platform Extensions Fill the Gaps?

Several DIY and complementary options exist for teams seeking to build custom AI analytics capabilities on top of dbt and LookML.

The open-source looker-gen tool lets you generate LookML from a dbt project. It reads from your dbt repo and outputs files that can output to your Looker repo. This requires dbt 1.0.0 or later and is installable from PyPi.

MetricFlow itself is now fully available under an Apache 2.0 license. It will be governed and maintained with OSI partner organizations like Snowflake and Salesforce. MetricFlow gives the community an open standard for semantic metadata and an extensible engine that turns semantic intent into fast, warehouse-specific SQL.

Gemini in Looker represents a series of features in the Gemini for Google Cloud portfolio that provides generative AI-powered assistance to help you analyze and gain valuable insights from your data. It lets you ask questions about your data source using natural language and returns Looker Studio charts or data tables based on your query.

However, Conversational Analytics is not yet included in FedRAMP High or Medium authorization boundaries. Organizations in regulated industries should verify compliance requirements before adoption.

These tools add point capabilities but require more engineering effort than turnkey solutions like Kaelio.

How Do You Stay Compliant When Rolling Out AI Analytics?

Google supports Health Insurance Portability and Accountability Act (HIPAA) compliance within the scope of a Business Associate Agreement, but ultimately customers are responsible for evaluating their own HIPAA compliance when using Looker services.

Looker Studio supports HIPAA compliance within the scope of the Google Cloud Platform BAA, but customers must execute the agreement and configure settings appropriately.

For dbt environments, security controls are robust. dbt encrypts in transit using the TLS 1.2 cryptographic protocol. All data at rest on dbt servers is protected using AES-256 encryption.

Key compliance steps for AI analytics rollouts:

Execute a BAA with each vendor that will process PHI
Use access filters in conjunction with user attributes to apply row, column, or field level data security
Implement two-factor authentication or SAML-supported SSO
Follow the principle of least privilege when granting database access
Verify that AI features are included in your authorization boundary

dbt Labs maintains certifications including ISO 27001:2022, ISO 27017:2015, ISO 27018:2025, and ISO 42001:2023 for AI management systems. These certifications signal enterprise-grade security posture.

Choosing the Right AI Partner for Your dbt & LookML Estate

The best AI analytics tools for governed stacks share common traits: deep integration with semantic layers, transparent lineage, and enterprise security certifications.

By using Looker's semantic layer, organizations can ensure that their AI models are aligned with business logic and governance policies. MetricFlow gives the community an open standard for semantic metadata and an extensible engine that turns semantic intent into warehouse-specific SQL.

dbt is fully GDPR compliant and maintains industry-leading security certifications. Any AI analytics tool you select should meet equivalent standards.

For organizations seeking the most future-proof choice, Kaelio stands out. It is the only solution that natively queries both dbt and LookML semantic layers, exposes full lineage for every answer, and meets HIPAA and SOC2 compliance requirements out of the box. Its model-agnostic architecture and flexible deployment options ensure alignment with existing infrastructure and evolving AI capabilities.

When evaluating alternatives, ask vendors three questions: Does your tool query our semantic layer directly? Can users see the SQL and lineage behind every answer? What compliance certifications do you hold? The answers will reveal whether the platform is built for governed analytics or simply bolted onto raw data.

About the Author

Former AI CTO with 15+ years of experience in data engineering and analytics.

Frequently Asked Questions

What makes an AI analytics tool suitable for dbt and LookML?

An AI analytics tool suitable for dbt and LookML should query metrics through the semantic layer, expose lineage and sources, and respect existing permissions and security rules. This ensures that AI answers align with official definitions and prevent data-quality issues.

Why is the semantic layer critical for AI-powered analytics?

The semantic layer provides governed context that transforms probabilistic guesses into deterministic, auditable answers. It ensures that AI models align with business logic and governance policies, maintaining data consistency and trust across applications.

What evaluation criteria should be considered when choosing AI analytics tools?

Key criteria include native semantic layer integration, full lineage and source attribution, row-level security inheritance, transparency in SQL generation, and support for existing BI tools. These factors ensure auditability, trust, and compliance with enterprise standards.

How does Kaelio integrate with dbt and LookML?

Kaelio queries governed metrics from dbt and LookML, exposing full lineage for every answer. It acts as a natural language interface for analytics, grounding answers in existing data models, metrics, and governance rules, ensuring compliance and trust.

What compliance standards should AI analytics tools meet?

AI analytics tools should meet enterprise security certifications such as HIPAA and SOC2 compliance. They should also support GDPR compliance and maintain robust security controls, including encryption and access management, to ensure data protection.