Fabi is the only AI data analysis platform that is truly accessible to the entire team. If we can summarize the theme in January: Continuing to create a self-service analytics experience for all, with improvements all the way from small UI features down to the data.

Application connector enhancements

Our Fabi 2.0 launch was a big step forward. We introduced hundreds of application connectors, making us the only AI data analysis platform that can analyze your data wherever it lives. No migration, no copying data around.

Since launch, we've made two improvements based on what we're seeing from customers:

Automated data modeling for additional sources. For certain connectors like QuickBooks and Stripe, we've added dbt models that automatically transform your data to make it as AI-ready as possible. What does this mean in simpler terms? Analyst Agent can understand your data better right out of the gate: more accurate insights with zero manual setup on your end.

More transparency on usage and billing. In your data connector settings page, you can see exactly how many records have synced by source for the current month. You can also adjust spending limits to make sure you don't exceed your budget if there's a sudden unexpected surge in record updates.

Tip: If you're using an application connector in Fabi and think you could be getting more from Analyst Agent, reach out. We're adding new models all the time and we're happy to offer complimentary data engineering services to our Teams and Enterprise customers.

An AI that uses its brain…

...or whatever it is under the hood.

Rather than systematically attempting to respond to every prompt immediately, Analyst Agent will sometimes ask clarifying questions when it thinks the ask is ambiguous. Not only does this help you think through the specifics of what you need, but it ensures that the response you get is much more relevant to what you're actually trying to accomplish.

Quality of life improvements

You may have already noticed a few new features we've rolled out to make sure we continue providing the most powerful and safe analysis environment:

• Table of contents in Smartbooks: Easily navigate your analysis

• Code variable explorer: For the more technical crew, you can see all the variables in a given Smartbook and quickly jump to the code block that generated it.

• Broken dependency warnings: If you deleted a block or cell that is used somewhere else, Fabi will warn you to ensure that you don’t accidentally break your analysis. These are the types of guardrails that we truly believe contribute to making analytics self-service.

  

Other interesting reads and content

Why legacy BI isn't working for AI

Marc wrote a guest post for Atlan's Metadata Weekly exploring why so many teams are struggling to get AI working with their existing BI stack—and what the teams figuring it out are doing differently.

Here are a few of the core reasons:

Legacy BI could work with poorly modeled data. You could patch things up at the BI layer, which just compounds the problem over time. Good data modeling is the #1 driver for AI accuracy, and traditional tools weren't designed to enforce it.

Traditional solutions operate on a rigid semantic layer paradigm. For drag-and-drop functionality to work with SQL compilers, you needed these strict layers. But that's not the case anymore. AI actually works best with role-dependent and plain English context. Smaller organizations are finding they're more comfortable having context live in .md files rather than complex proprietary schemas.

A lot of teams try to nail self-service analytics for everyone in the organization rather than taking incremental steps. Most successful organizations start with augmentation of data-literate teams and cut their teeth on these use cases before trying to deploy an AI analyst that can answer every question for the business.

Small and medium organizations are starting to figure out what large enterprises are struggling with. This is natural—the shift to AI-first requires a shift in strategy and thinking, and it's dramatically easier to make this shift when there isn't a deep-rooted foundation to work around.

Looking at what these organizations are doing should give data leaders some good clues of things to come.

Read the full post: The AI Analyst Hype Cycle

Building an MCP server?

We were building our MCP server at the same time as MotherDuck was building theirs, and we swapped some notes. They did a nice job touching on some of the more tedious technical details, with a particular focus on authentication: motherduck.com/blog/dev-diary-building-mcp

This is a must-read for data and engineering teams dabbling in your own MCP server build.

Tip: Interested in an AI data analyst MCP server but don’t want to build one from scratch? We’ve got you, check out our Analyst Agent MCP server.

​​Today we’re introducing Fabi 2.0 - The first AI data analyst that can speak to all your data

We founded Fabi with a simple but bold vision: Getting insights from your data should be accessible to all and should be a team sport.

Over the past year, since launching the first version of Fabi, we’ve gathered a tremendous amount of feedback and have worked with product, growth marketing, data and many other teams to deliver the world’s best data analysis experience.

With Fabi 2.0 we’re officially making a number of features generally available:

• Hundreds of applications connectors from Stripe and HubSpot to Shopify and TikTok

• A simple AI configuration interface to give context about your business in plain English

• Our MCP server so you can talk to your data wherever you like and embed agents right into your workflow

Watch our full announcement and support us on LinkedIn with a like and comment.

Or if you’re more of an X/Twitter user, show some love there.

Thank you for being on this journey with us. We’re just getting started.

Marc, Lei and the rest of the Fabi team.

​​We started with a simple vision: getting insights from your data should feel as natural as asking a colleague a question. No waiting on engineering tickets. No SQL knowledge required. Just you, your data, and AI that understands what you need.

In 2025, we proved this wasn't just possible, but what teams have been waiting for.

Fabi’s general availability marked the starting line and your feedback shaped every feature we built:

What you told us you needed:

• Control over AI behavior and model selection

• Confidence to experiment without breaking things

• Ways to share insights across teams

• Integration with tools you already use

• Speed without sacrificing accuracy

What we built:

• AI Analyst Agent that learns your business context

• 100+ database and data source connectors

• Features for technical teams: Typeahead, GitHub integration

• Enterprise-grade collaboration features

• MCP server for chat-based data access anywhere

The numbers tell the story: 92% faster analysis at companies like Aisle. Hours of manual work eliminated. Non-technical teams answering their own data questions.

📊 2025 by the numbers

Everything we do is for you, our users. By signing up for Fabi you’ve decided to join thousands of other teams looking for a better way to analyze their data. We wanted to share some fun numbers with you:

That’s a whole lot of insights delivered! As for how much AI you all have used, for Claude alone: 

Ultimately, the best proof that we’re on to something real is what our users have to say:

🤖 AI Analyst Agent: Built for accuracy at scale

We built our AI Analyst Agent from the ground up using an agentic framework with tool calling. The result? Analysis that's 10X more accurate and contextually aware.

Our technological breakthrough:

Business context that sticks: Teach the agent your terminology and data conventions once. Set org-wide context so everyone gets consistent, accurate answers.

Tool calling for precision: The AI queries multiple sources, validates its own work, and handles multi-step analysis without breaking a sweat.

Save and share your work: Ask questions in plain English, refine through dialogue, build dashboards without code. The seamless flow between conversation and execution is unlike anything else.

What our customers love:

Teams using Fabi AI Agent complete analysis in minutes rather than hours. Complex queries that once required multiple back-and-forths now work on the first attempt.

🔁 Deliver insights where your team works

And finally, we also shipped a new way to deliver and collaborate around insights.

What we built: AI data workflows that run on your schedule. Set up analysis once, deliver insights automatically.

Recurring analysis delivered: Workflows enabled teams to generate reports on a regular cadence and send them to their teams via Slack, email, etc. 

Alerts for what matters: Create thresholds for sensitive metrics you care about once and get automatic alerts when they change. 

Executive-ready insights: Auto-generated executive summaries for analysis that cut report-building time and create insightful visualizations.  

Teams went from spending hours on recurring reports to spending minutes setting up automation. An analysis that used to require manual effort every week now runs while you sleep.

🚀 What's Next

We're not slowing down. 2026 will bring deeper integrations, more powerful automation, and AI capabilities we haven't announced yet.

Thank you for building with us this year. Your feedback shaped every feature we shipped and we look forward to more to come in the new year.

Happy holidays,

Marc, Lei and the rest of the Fabi team

🔌 Application connectors: Connect to all your data

From day one, our goal has been simple: help teams find and share insights incredibly fast, no matter where it lives. Today, we're making that vision a reality with our new application connectors. Most companies have data scattered everywhere: databases, CRM tools, product analytics platforms, marketing systems, and dozens of other apps. Now you can connect them all, from Amplitude to Stripe and everything in between.

This is our biggest unlock to date and it’s available right now.

We're excited to see what you'll build in Fabi with all your data finally connected.

Pricing details: Application connectors are part of the free trial and start at $100/mo afterwards.

🏠 Redesigned home experience: From question to insight in seconds

Fabi is the fastest way to do any analysis and we strive to remove as many clicks as possible. We've streamlined how you kick off an analysis in Fabi to eliminate the friction between having a question and finding an answer.

What's new:

• Zero clutter: A clean, focused interface that gets out of your way

• One-click start: Go from question to active Smartbook instantly

• Faster discovery: Find recent analyses, templates, and data sources without hunting

The result? You spend less time navigating menus and more time exploring your data. Whether you're building a new dashboard, running an ad hoc query, or picking up where you left off, the new home experience gets you there faster.

🎙️ Podcast

Data science is here to stay

In the recent episode of the Data Neighbor Podcast, Lei, one of our co-founders and CTO, shares why the future of data science isn't about who writes the best code but about who understands the business and asks the right questions. As AI transforms how we work with data, the human skills of curiosity, context, and critical thinking become more valuable than ever.

Why AI needs guardrails to be useful

Lei also joined a conversation about building AI products on Timothé Frin's documentary series. The episode features insights from leaders at OpenAI, Google DeepMind, YC, and more, all discussing how AI is reshaping how we build products. Lei's key insight: AI tends to "boil the ocean." Without explicit guardrails, it will try to refactor everything instead of solving the specific problem at hand. This speaks to something we think about constantly at Fabi: how do you give AI enough freedom to be powerful, but enough structure to stay focused?

🎥 New tutorial: Chat with your data from Claude in under 10 minutes

Stop waiting for dashboards or writing SQL queries. This step-by-step video tutorial shows you how to connect any database to Fabi's AI Analyst Agent and integrate it with Claude Desktop via MCP for instant ad hoc analytics.

What you'll learn:

• Connect your database (we use Supabase in the demo) to Fabi in minutes

• Set up the Fabi MCP server with Claude Desktop or Cursor

• Ask questions in plain English and get instant answers from your data

• Extend the setup to Slack, email, or other tools your team already uses

Perfect for founders who need quick metrics, product managers exploring user behavior, or anyone who wants to move faster without learning SQL.

📺 Live demo: Automating your data workflow

Most teams drown in data but starve for insights. Workflows solve this by sending the right information to the right people at the right time.

In this hands-on session, we'll build three real workflows together:

• Dashboard vs. workflow: Learn when to visualize vs. automate

• Workflow anatomy: Understand how data flows from source to action

• Multi-channel delivery: Push insights to Google Sheets, Slack, and email

• Live builds: Watch real workflows come together with best practices

This session is for product managers monitoring metrics, sales managers tracking pipeline health, or founders keeping tabs on customer acquisition. You'll see exactly how to automate insights that move your business forward.

Get more control over data context, more natural conversations, and visual enhancements that make every dashboard shine. We’ve improved how you explore, analyze, and present your data so that you can move from question to insight with zero friction.

🛠️ Enhanced data source management

We've redesigned how you manage and configure your data sources to give you greater control and clarity. With our new data configuration panel, you can:

• Edit and rename data sources directly from any Smartbook

• Add context for the AI through plain-English descriptions that help the AI understand your unique business terminology and internal data conventions

• Refresh schemas instantly to stay in sync with your live data as it evolves

Data source configuration

⭐️ Customize Analyst Agent

You can control your AI experience with our new Analyst Agent Settings panel to:

1. Default AI model: Pick your preferred AI model from your favorite provider

2. Longer AI memory: Adjust memory retention length for more accurate responses

3. Custom prompt: Include a custom prompt to adjust the AI to your needs. Give your brand colors, tell it how verbose to be or your preferred charting library.

Analyst Agent chat settings

🧠 Extended conversation memory for Analyst Agent

The AI Analyst Agent now retains more historical context throughout your conversation until you start a new thread. This enhancement ensures the AI maintains awareness of your entire conversation journey, making each interaction more informed and productive.

Note: This is available to customers on the Builder plan and above.

🌈 Better visualizations with Altair

By adopting Altair as our default visualization tool, we've dramatically enhanced the clarity and aesthetic quality of every chart you create, delivering a more polished, professional look across all your data visualizations.

Chart made with Altair

📹 New videos

🗞️ Latest blogs

🚀 The next big update is almost here

We’re working on some game-changing analytics features.

Reach out to us if you’re interested in beta-testing new features we release, and make sure you follow us across all our social media channels to hear about new features and products.

1. Subscribe to our newsletter: https://www.fabi.ai/blog 

2. Product Hunt: https://www.producthunt.com/products/fabi-ai 

3. LinkedIn: https://www.linkedin.com/company/fabi-ai/ 

4. X: https://x.com/hqfabi 

📄 Focus without distractions

Sometimes, you need quick, conversational answers. Other times, you need structured workflows. Now Fabi gives you both.

With our full-screen chat mode, you can enjoy a fully immersive chat experience, ideal for asking quick questions, testing ideas, and saving your work as a repeatable analysis. When you're ready, you can easily transition from conversational analysis to structured Smartbooks without any compromises or silos.

Behind the scenes, our engineering team has achieved a major technical breakthrough that enables the management of multiple kernels and tracks variables without conflict, making the transition from exploration to collaboration smoother and more powerful than ever. And if this is more than you understand, fret not, we’ve made sure everything works like magic under the hood.

🧠 GitHub integration for technical collaboration

Vibe analytics doesn’t have to mean compromising on engineering best practices. With our GitHub integration, version control is a breeze. Technical data practitioners can collaborate effectively and Fabi can be fully integrated into CI/CD pipelines and automations.

You can read more about our GitHub integration in our documentation.

🏎 New connectors

Our goal is to empower you to analyze your data seamlessly, wherever it lives, and we're committed to delivering on it. Here’s what we added last month.

• Coda: Pull in your Coda tables by simply copy/pasting the URL and letting the AI do the rest.

• Firebolt: Analyze your massive production datasets right in Fabi.

  

⭐️ Seamlessly add data to your AI analyst context

Part of working with AI is understanding the context it’s using, and we also wanted to make it even easier to add data directly from a single location.

We’ve added the data source context right in the AI Analyst Agent interface, giving you a quick glance of the data available to you and the ability to add data from any of your data sources: Google Sheets, databases, Airtable…

🎁 Other goodies

• Experience faster, more responsive data apps with optimized filter and input caching

• The latest and greatest Anthropic model Claude 4.5 is now available in Fabi

• We’ve made some improvements to Excel and Google Sheet handling so that the AI can better navigate complex spreadsheets with merged cells, multiple headers and all the other fun stuff

The world’s most powerful vibe analysis platform just got even more powerful 💥

In August we continued to push hard on our vision to give you the most amazing data analysis experience so that you can get insights from your data faster, build data apps faster and automate your work with AI.

🪄 Auto-accept AI suggestions

We noticed that more than 95% of the code run in Fabi is AI generated. What this means is that you’re spending most of your time just talking to the AI and accepting its suggestions. Given this, we figured that we might as well save you a step, and let you just auto-accept AI suggestions as you build out your workflows and dashboards.

We’ve introduced AI “auto-accept” mode which does just that! When you’re running the AI Analyst Agent in this mode, it will figure out the right code blocks to update for you and just do it.

⏮️ Version history

Part of vibe analyzing is knowing that you can interact with AI with the full confidence that you can revert any change. Last month we introduced the Revert functionality, which allows you to quickly revert any changes you’ve accepted.

This month we’re introducing broader version capture both at the Smartbook and at the cell level. And here’s the coolest part: We snapshot everything for you automatically at key moments.

Think of this as checkpoints when playing your favorite video game 🎮. No more need to remember to save your progress at the risk of your parents unplugging the console!

📃 Improved file and Google Sheet management

Two of our most popular data sources: Google Sheets and uploaded files.

We wanted to make it even easier for you to just drop a file into Fabi and use our AI to analyze it. So here are some improvements that we shipped:

1. In Google Sheets cells, you no longer have to specify the sheet you want to pull. We will automatically connect to all sheets in a given Google Sheets

2. When you select a Google Sheet we’ll auto run the cell for you to import your data

3. We’ve improved the variety and complexity of files. For instance, if you upload an Excel files with multiple sheets that have pivot tables, charts or merged cells, our AI is able to better handle those cases

🎁 Other goodies

• More connectors: We’ve added support for Supabase, Neon and MariaDB

• GPT-5: We now offer access to all the latest GPT-5 models

• Suggested questions: The AI analyst agent will now suggest follow-up questions. This is great when you’re just starting out exploring a new dataset!

Follow us for more

If you’re as excited about the future of AI for data teams as we are, here are a few places where you can stay up to date:

• YouTube channel

• Fabi‘s LinkedIn and Marc’s LinkedIn

• X (formerly Twitter)

The theme for this month: Providing the best quality-of-life experience for you. You’re building so much and so fast with AI that we’ve received feedback that you want help keep things clean and organized.

⏪ Revert AI changes

Vibe analytics is all fun and games until you realize 5 messages deep that you took a wrong turn (it happens to the best of us). We saw this as being an inhibitor, preventing you from moving as quickly as you would if you knew that you could quickly revert any mistakes.

With the latest update, we’ve introduced a revert button right in the AI Analyst Agent interface. Now you can quickly scroll back up and revert the Smartbook to whatever point in time you’d like.

Revert your Smartbook to any previous version

🗂️ Organize your Smartbooks and dashboards

The number of analyses, dashboards and workflows we’ve seen our users creating has absolutely exploded. We’ve received many requests to help keep your work organized, so we’ve introduced folders and workspace sections in the navigation menu.

And as an added bonus we’ve added more granular Smartbook and dashboard sharing permissions. You can share these with a specific set of individuals.

• Favorites: Your starred Smartbooks, dashboards and workflows

• Organization: Items shared with the entire organization

• Shared: Items shared with a subset of individuals

• Private: Items only available to you

Workspace organization

✨ AI model selector

In the spirit of always letting our users choose the AI models they want to work with, we’ve added the option to select the AI model to use in our AI enrichment and AI summarization cells.

Now you can test out different models and balance speed and accuracy to fit your needs.

Select your AI model in AI cells

🎁 Other goodies

A few more fun things for you:

• Improved AI actions suggestions: With better dependency tracking, it’s now easier than ever to vibe analyze. The TL;DR is that you can jam as much as you want with the AI without accepting/rejecting changes until you get the final results you want, and it will all work perfectly!

• Smartbook status indicator: In the left hand menu and in Smartbooks we’ve added some indicators to tell you at a glance if an analysis was published as a dashboard, if it was scheduled and if the latest work you’ve done in your Smartbook is out of sync with the dashboard you published.

Follow us for more

If you’re as excited about the future of AI for data teams as we are, here are a few places where you can stay up to date:

• YouTube channel

• Fabi‘s LinkedIn and Marc’s LinkedIn

• X (formerly Twitter)

Over the past few weeks we’ve been sharing updates about our data analysis workflow functionality and today we’re officially launching it to the public. The feedback has been incredible and we’re really excited to bring an entirely new way of doing data analysis and distributing insights in the enterprise. Soon enough, we’ll wonder how we worked with clunky static dashboards in legacy BI.

🔀 Fabi Workflows Generally Available

Fabi Workflows are a completely new way to create and share insights. Not only can you use SQL and Python for data analysis (of course with the help of AI to vibe code), but now you can insert AI as an actual step in your workflow.

Say you want to analyze opportunity notes fields from CRM data stored in Snowflake and send out daily executive summaries in Slack. You can automate this workflow in no time.

Check out Marc’s announcement on LinkedIn.

Here’s a recap of key functionality:

• Pull data from your data warehouse, Google Sheets or Airtable (and let us know if you’re interested in other sources!)

• Analyze data with the help of AI, including batch processing plain English fields in your structured table

• Push charts, tables and AI-generated summaries to Google Sheets, email or Slack

• A new way to view your analysis with our workflow canvas, showing you all the links between your analysis components

🖼️ Template gallery

Fabi is actually becoming so powerful, that we’ve gotten the feedback that it can be hard to even know what’s possible on the platform! This is a great problem to have, but in the spirit of providing the most delightful data analysis experience, we thought we could make this easier.

We’ve published a template gallery to help inspire you. Click on any template to see the details and try it out for yourself in Fabi.

Here are some of our more popular templates:

• Building dashboards on Google Sheets and using AI to “talk to your spreadsheet”

• Performing sentiment analysis on natural language fields

• Talk to your dashboard with embedded AI

🔌 Databricks connector

Databricks users can now also enjoy all the power of Fabi for data analysis!

We add new integrations as demand for them comes in. If you don’t see a data source you need, please reach out to us at hello@fabi.ai and we’ll be happy to work with you.

🎁 Other goodies

A few more fun things for you:

• OpenAI o3 is now available: o3 has been added as a model option

• Documentation integration in AI Analyst Agent: Our AI is now integrated with our documentation, so if you hit a roadblock it can redirect you in the right direction

• Improvements to the AI: We made some big updates under the hood to how the AI works. We performed some context engineering and added a few new tools to continue providing the world’s best AI data analysis experience

Follow us for more

If you’re as excited about the future of AI for data teams as we are, here are a few places where you can stay up to date:

• YouTube channel

• Fabi‘s LinkedIn and Marc’s LinkedIn

• X (formerly Twitter)

We’re witnessing a shift in the way data teams operate. In the past data teams would spend most of their energy cranking out reports, leaving it up to the viewer to extract insights. Now data teams are bringing insights to their stakeholders in their favorite workspaces with AI playing a key role in generating insights.

In the previous product updates we introduced the new workflow canvas view and Push to Slack and Send Email cells. In this release, we’re introducing ways to insert AI as part of your automated workflow along with improvements to the workflow canvas.

✨ AI summarization and enrichment cells

AI has been transformative for code generation for data analysis, but what about actually using LLMs as a step in the analysis process?

For example, AI is perfectly suited for:

• Sentiment analysis

• Intent categorization

• Summarizing data

To this end, we’ve introduced two new AI cells:

1. AI enrichment cells: You can pass a structured dataset to the AI with a prompt that will apply to a designated field. If you have a messy text field, this cell can quickly make sense of it and produce a new field in that table, with no risk of hallucination on the original data.

2. AI summarization cells: Want to send a daily executive summary of your data to Slack or email? No problem, give the AI summarization cell your data with a prompt and then use that output to send the summary off to your stakeholders.

AI cells allow you to embed AI as part of your workflow

🔑 Secret manager (Vault)

Secrets secrets are no fun, when they’re shared with everyone.

If you’re using Fabi to automate workflows and dashboards involving third-party applications that require authentication, it’s really important to handle those secrets with care. We’ve introduced a Secret Manager for you to securely store those secrets and keep them safe.

Important note: At this time, secrets stored in Fabi are available for use by any Builder. We plan to introduce personal secret management. Please contact us if you would like us to consider prioritizing this work.

Fabi’s Secret Manager

Go to your Secret Manager, or learn more about how it works.

😍 Improvements to workflow canvas

The workflow canvas view is a fantastic way to see how everything connects together in your analysis. We spent some time making the user experience even more delightful.

• Scrolling around and zooming in and out is much easier (to zoom in click Cmd + and scroll with two fingers)

• It’s easier to highlight lines of code without moving the cell placement

• The list of variables and Python DataFrames produced by each cell are listed under the cell

You can see the workflow view (and the new AI cells!) in action in this video.

🎁 Other goodies

But wait, there’s more!

• Claude 4 is now available (for paying customers)

• Easier access “Run” button on the left hand side of any Smartbook or Workflow

• You can copy/paste a table up to 200 records by right clicking on the table

Follow us for more

If you’re as excited about the future of AI for data teams as we are, here are a few places where you can stay up to date:

• YouTube channel

• Fabi'‘s LinkedIn and Marc’s LinkedIn

• X (formerly Twitter)

After talking to literally hundreds of data teams and data practitioners, it became clear to us that you were trying to do more with your data than just build reports. You want your data and insights to be automated and shared, which requires that they be distributed at the right place and at the right time. And there’s no better place to meet your stakeholders than where they live every day: email, Google Sheets and Slack.

You can already push DataFrames to Google Sheets, which many of our users already love and use, and this month we’re introducing email and Slack cells.

✉️ Finally a better way to send reports via email and Slack

We’ve talked to a number of teams about their email and Slack reports, and a few things jumped out at us:

→ They’re mostly just screenshots that are illegible and crammed

→ There’s no way to easily update the content dynamically based on conditions

→ They’re hard to set up and maintain

With email and Slack cells, we’ve made it easier than ever to construct emails and Slack messages where you get to choose what Python variables, DataFrames and images appear. Simply write an email like you normally would, and insert your Python variables and objects using templating:

Building an email using Python and templating

How-to tutorial video.

🎮 Multi-player mode

With more and more users coming together in Fabi.ai Smartbooks to collaborate around insights, the feedback was clear: you wanted a way to collaborate on the same Smartbook.

Now with a click of a button you can give edit rights to your coworkers, and any conflicting changes will be flagged to make sure you’re not stepping on each other’s toes.

Sharing Smartbooks with coworkers

📡 Sharing report externally

From startups sharing reports with customers to marketing agencies building performance dashboards for their clients, there’s a clear need to share Fabi.ai Smart Reports with individuals outside your organization. Now you can do just that.

Sharing Smart Reports with clients

🫣 Sneak-peek of things to come

Between the new Workflow canvas view and the new cells that we just announced, you might be starting to see a theme emerge. We have a lot more coming on this front, giving data practitioners a completely new way to build insight workflows like you’ve never seen before. If you’re interested in taking a peek into the future, we’re always happy to chat!

We’re constantly amazed and inspired by what you build in Fabi.ai. We’re feeding off that creativity to make it easier and faster for you to get the insights you want from your data and to share those with your team.

🔀 Workflow view

We’ve had many conversations with our community, and we’ve learned a few things:

• Our users are creative and you started using Fabi.ai in ways that we hadn’t imagined: automating workflows and reports

• You’re not satisfied with Slack and email reporting capabilities in your existing solutions

• You want do do even more with your data like natural language processing and document processing

We’re working on some very exciting features that we’ll share with you very soon (or reach out if you’d like a sneak peek), and this Workflow view is a taste of things to come. For now this Workflow canvas is simply a way to visualize the workflows you’re already building, but there’s more to come.

✨ Choose your favorite AI model

We’re committed to remaining LLM-agnostic, and as part of that, we now let you pick which model you want to use!

We’re continuously adding new models as they get released.

Tip: We’ve tuned Fabi.ai for the most popular models and Claude 3.5 Sonnet and 3.7 Sonnet remaining the highest performing in our opinions, but if you would like us to tune a specific model just reach out!

Titles and descriptions in Smart Reports

We’ve always believed that the way insights are communicated is just as important as the actual insights themselves. To help improve collaboration between data practitioners and their stakeholders, we’ve added the option to add titles and description to all tables and charts in a published report.

Hey there, {{ First name | friend }}! 👋

Today, we're excited to announce Analyst Agent - a new way to make self-service analytics a reality with specialized AI data agents that your team can trust. ✨ 

What is Analyst Agent?

We've reimagined how AI and analytics work together. Instead of connecting a general AI to your data warehouse and hoping for the best, Analyst Agent lets data teams deploy specialized AI agents that work exclusively with curated datasets - ensuring reliable insights while maintaining complete control over data quality and access.

Analyst Agent can answer questions directly in Smart Reports.

What makes Analyst Agent different?

• Deploy focused agents in minutes, not months

• Enable Python-powered analysis without writing code

• Maintain nearly 100% accuracy within defined domains

• Keep full control over data access and quality

Key features:

• Dataset-based architecture - Create datasets using Python DataFrames

• Real-time data syncs - Curated datasets stay in sync with underlying data sources

• Multi-step validation - AI that checks its own work

• Universal data connectivity - Connect to any data source

You can learn more about what went into our Analyst Agent build (and how you can use it) in the full write up over on the Fabi.ai blog. You can also find the full demo and docs here. 

Supporting us on Product Hunt ♥️

We'd love your support! Upvote Analyst Agent on Product Hunt and share why you're excited about specialized AI data agents.

Best,
Marc @ Fabi.ai 

2025 is the year of AI agents, and we’re bringing a new, game-changing type of agent to data and analytics teams.

In this update, we’re sharing a preview into things to come with early access of our latest feature for our community (that’s you!)

🤖 Analyst Agent - Early Access

Since mid-2024 we’ve been working on a new form of analyst agent to support data practitioners and it’s been enthusiastically adopted and helps cut down analysis workloads by up to 94%!

But our vision has always been to bring these AI capabilities to the business user as well. This helps remove some of the load on the data team, while creating a data-driven culture. But it was critical to us that the data team be the one to build these agents so that you can trust them as you deploy these.

Now in Fabi.ai Reports, AI is automatically embedded and the AI agent is trained to answer any questions about that specific report. This is powerful because:

• Self-service analytics: Business users can explore questions about the data on their own instead of following up with additional questions.

• Data teams controls the agent configuration: These AI agents are trained to only answer questions on that report. No risk of them going rogue and creating random SQL queries that mislead the team.

• More than just SQL: Under the hood the AI agent operates in Python. Why does that matter? Because Python can do more than just slice and dice data. You can ask the AI to figure out which fields are leading indicators, or run a regression analysis to score leads or even provide a time-series forecast.

Check out a demo walk-through here (5 min) with some tips on how to configure your AI agent.

Analyst Agent in Fabi.ai reports can handle follow-up questions for you.

✨ Other goodies

• New help center and videos: From any Smartbook you can easily access our help hub by clicking on the question mark. We also have a refreshed YouTube channel with lots of tips and tutorials.

• Bugs fixes and improvements: We’ve addressed some issues with Excel file extensions and database schema crawling

To start off the new year, we wanted to share our latest features, some fan-favorite features from 2024 and a sneak peak of things to come.

🎨 Table formatting with AI

Anyone who has used a spreadsheet knows how important conditional formatting can be. So we’ve brought that functionality right to our tables and juiced it up with AI. Now you can ask the AI to add whatever conditional formatting you can dream up, and you no longer need to fiddle around with a million dropdowns.

Conditional formatting using AI

🔌 New integrations: MotherDuck and ClickHouse

By popular demand, we’re introducing two new connectors:

1. MotherDuck: A high-performance cloud data warehouse powered by DuckDB. Great for building data apps or conducting analysis on small and medium datasets.

2. ClickHouse: Another high-performance, column-oriented SQL database management designed specifically for analytical purposes.

Learn more about connecting to your data warehouse in our documentation. And remember, with Fabi.ai you can connect and query multiple data sources in a single Smartbook!

Highlights from 2024

With 2024 in the books, this is a great opportunity to reflect on the powerful new features we’ve added to our AI data analysis platform:

• AI agent for data analysis: Our new AI agent designed specifically to boost productivity of data teams and data practitioners was made generally available. Our AI can now use tools to auto-install Python packages, dry-run code, retrieve historical context and much more.

• Flexible layouts for powerful report building: Turning an advanced data analysis into a beautiful, interactive report powered by SQL and Python has never been easier. Now with our drag-and-drop report building and filters, you can collaborate more efficiently with your stakeholders.

• Python to Google Sheets integration: Your stakeholders just want a spreadsheet? No problem! You can retrieve, clean and analyze data with SQL, Python and AI and push DataFrames directly to Google Sheets and even schedule updates.

What’s next for 2025?

Our mission to build the world’s most powerful collaborative data analysis platform continues. You’ll see a few big themes coming from us this year:

• More building blocks: Every organization has unique data and unique needs. The Fabi.ai collaborative data analysis platform is your canvas to bring your vision to life. You can expect more integrations in and out of Fabi.ai and more tools to analyze your data.

• World-class data analysis AI agent: We’re going to continue investing in our AI agent toolkit to continue providing the world’s most powerful AI data analysis agent. Our AI will be able to interact with your data in more ways than ever.

• Low-code tooling: We love SQL and Python, but bringing data analysis to the masses requires low and no-code tooling. We’re going to continue simplifying the data analysis process, leaving the coding to the AI or technical users.

Our mission is to make it incredibly easy for you to conduct data analysis and share your insights with your coworkers as beautiful, timely reports. With our latest major release of “flexible reports layouts”, we’ve taken yet another major step in that direction.

🧑‍🎨 Customize your report layouts

Building a beautiful, interactive report that looks just the way you imagined it, with customized visualizations has never been easier. Now with our flexible reports layouts, you can remove, resize and reorder tables, charts and filter in seconds.

Check out our documentation for an explainer video and more details.

🤖 AI Agent (Generally Available)

Not just another buzzword. AI agents are a clever way to improve speed and accuracy.

How does it work? Give the AI an objective and some tools (functions) to complete that task and let it work through the task. The first iteration of our AI agents auto-tests the code and can fix multiple code issues at once, helping you get to your answers faster. This is perfect for more reliable AI outputs and AI auto-debug. No more “That resolved my first error, now I’m getting a new error…”

Our AI agent will also auto-install Python packages if an imported package is missing.

⏰ 100% customizable report schedules

Want to schedule your report or Slack alerts to run every other weekday at 5AM and 5PM unless it’s a full moon? No problem. Well almost… cron jobs don’t yet track lunar cycles. Just ask the AI and it will write the cron expression for you so you don’t have to.

✨ Other goodies

• Check out our article and video explaining how to quickly build a Slack app to send reports and custom notifications

• Our tutorial on how to create a Sankey diagram with Python and AI is one of our most popular pieces of content. Go check it out and build your own in minutes!