Current state of AI Analytics
If we get to the bottom of things, all analytics tools have enabled just one thing: to get from raw data to insights that can be interpreted by non-technical humans and acted upon to drive business outcomes. AI analytics leverages the emerging AI technology at certain stages in the pipeline to primarily deliver on two improvements:
Make analytics faster for data analysts and engineers.
Allow a broader range of non-technical individuals to gain insights without the assistance of data engineers.
This list contains all the tools on the spectrum—from full-stack analytics platforms to specialized developer tools solving a specific problem.
AI analytics could grow massively, but it's still in its infancy at the moment. This is not surprising, given that the underlying technology is a living and breathing substance that is yet to shape into something that anybody can understand.
As always with emerging and fragmented markets, there are lots of AI analytics tools that pop up. All of them are trying to figure out the best ways to solve existing problems in new, unique ways. Oftentimes, their use cases overlap, and there are no clear winners at the moment. In our view, this is the perfect time to jump in and get to understand this landscape and leverage this emerging technology to get a head start.
In the face of emerging technology, it's very beneficial to get to the bottom of things and look at all of these tools from the perspective of the outcomes these tools deliver on or the "jobs" that users "hire" these tools to deliver.
Categorization
To make this directory useful and easy to navigate, it is essential to categorize the tools. Although categorization is somewhat arbitrary as startups add new features or make soft pivots toward other audiences, it can make it possible for them to migrate from one category to another.
| Text to SQL | The simplest of the bunch, this was one of the first categories to emerge after ChatGPT came out. It is generally seen as demo software that's not suitable for real-world use. |
| AI Analyst | The biggest common denominator for tools in this category is that (at least in their vision) they strive to take on the job of a full-time analyst so that every non-technical employee can get self-serve business insights without SQL/Python/data engineering help. This most often includes several "jobs": chatting with a natural-language LLM agent that runs queries against data sources under the hood; generating dashboards, reports, and data products; and distributing the insights across teams and channels. These tools are the natural evolution of the basic "text-to-SQL" tools, but unlike them, AI analyst tools more often than not require data engineers to correctly set them up on top of existing data infrastructure before business users can use them. |
| Semantic layer | Semantic-layer platforms map complex data into familiar business terms so that all users can access the same source of truth, with full confidence in its integrity through this unified enterprise data layer. It helps keep all definitions and business logic in one place and then manage and change them centrally. Typically, these tools provide the biggest benefits to larger enterprises that have lots of teams, each of which has its own definition of "gross revenue," so the costs of AI agents misinterpreting natural language pile up. Most of these tools provide natural-language AI agents, which makes them similar to the AI Analyst category; however, the latter has a much weaker focus on semantic modeling in favor of usability and speed. |
| AI Spreadsheets | Tools in this category, unlike AI Analyst tools, focus on spreadsheets/csvs as data sources and provide a spreadsheet-based interface. On top of it sits an AI chat that allows users to perform calculations, data cleaning, and create pivot tables using natural language. Some of the tools offer a plugin that brings AI chat and other functions directly into the web versions of Excel or Sheets. The most sophisticated tools in this group allow for extra flexibility with Python or JS cells and database/API connectors. |
| Data IDEs | Similar to traditional IDEs, data IDEs are AI data editors designed for data teams — connected to the data warehouse and business context, and powered by an agent that assists in writing SQL and Python code. They allow data scientists and engineers to create data pipelines, run analytics, explore data, and ensure data quality from a single integrated space. |
| Qualitative AI Analytics | AI qualitative analytics tools help teams analyze unstructured customer data like interviews, support tickets, sales calls, surveys, and open-text feedback at scale. These tools detect intents, emotions, pain points, and emerging patterns, turning raw conversations into structured insights. Some of these tools focus on concrete use cases like exploratory analysis for academia or extracting issues coming from user<>LLM interactions. Others focus on product analytics or sales interactions, providing additional capabilities like sales-team training. |
| Data Pipeline Assistant | Tools in this category use AI to help data teams design, build, and maintain data pipelines with far less manual effort. It understands warehouse schemas, lineage, and business logic, and automates repetitive tasks like mapping fields, generating tests, fixing broken queries, and detecting upstream/downstream impacts. These assistants reduce cognitive load, improve data quality, and speed up development cycles by acting as always-available copilots for data engineers. |
Expectations
Set Realistic Expectations
Given the state of the market, don't expect perfectly working apps; there's no beaten path here, so expect effectiveness gains at the expense of stability, scalability, and security.
Best practices using the directory
First, go over the categories to get an initial sense of what each category does.
Then, go to the list of tools and filter by categories you're interested in learning more about.
View the details of the tool by clicking on it. The section is structured using a jobs-to-be-done framework in mind:
What it is for
Listing the problems this tool addresses.
Who it is for
Listing the types of users this tool targets and who it solves the problems above for.
Benefits vs Status Quo
Providing an overview of how these problems were solved in the past and what benefits this new solution has compared to them.