Marc Dupuis

Fabi.ai Workflows - AI-powered data workflows with SQL + Python in one platform

With Fabi.ai workflows, data teams, founders and PMs can automate insights to and from their data warehouse, Google Sheets, Airtable, Slack and many other sources and destinations. Bring insights right to where your users work.

Add a comment

Replies

Best
Marc Dupuis
Hey Product Hunt! Super excited to launch [Fabi.ai](http://Fabi.ai) Workflows here today. We've been working with product and data teams across a number of organizations, and have seen the need to automate insight delivery to where stakeholders actually do work vs. just deliver more dashboards. At the same time, we know that automating that analysis and delivery has historically meant context switching between a ton of tools, often looking like: - Analyzing data in SQL/Python notebooks - Manually exporting results to spreadsheets - Copy-pasting insights into Slack updates - Writing custom scripts to send alerts to Slack, email, or back to Google sheets - Struggling to meaningfully integrate AI into the actual analysis step of the workflow We saw an opportunity to simplify this work and make it completely plug-and-play + add in AI functionality as part of the full data analysis workflow. This makes it easy to pull data from multiple sources (warehouse, Sheets, APIs, etc.), apply traditional analysis and AI capabilities (like sentiment analysis on customer reviews), and push those insights to stakeholders, all without leaving [Fabi.ai](http://Fabi.ai). 🔧 What we built: - Workflow Canvas View: Switch between our familiar notebook-like interface and a visual workflow canvas that gives you a Bird’s eye view of your workflow (even when it gets complicated). It’s great for collaboration and understanding complex analysis at a glance. - Workflow Cells (or Blocks): Every cell takes Python objects and variables as input and produces Python objects as outputs. We stick the lingua franca, making it super simple for anyone already doing data work to pick up and use. - Low-code Integration Cells: Our most popular ones are Google Sheets I/O cells (to pull or push data ), Airtable, and Slack. Just pass any variable, dataframe, or image using {{your_python_object}} syntax. The coolest part is that our AI can build and insert these cells for you automatically (though we’re still polishing this feature). - AI Analysis Cells: When you need sentiment analysis, AI summaries, or any text processing that Python isn’t the best tool for, use our AI enrichment or AI summarization cells. Give them a dataframe and a prompt and they’ll generate new columns or summaries based on your instructions. You can read the full details of what went into our decision to build Workflows (and how to use it) in our latest blog here: [blog link] You can check out some of our templates (including sentiment analysis reports, closed-lost deals summaries pushed to Google Sheets, and AI-generated summaries for key accounts in Slack) here to get a jump start: https://www.fabi.ai/template-gal... [Fabi.ai](http://Fabi.ai) Workflows are included in our free Starter plan, so you can build and schedule your first workflow for free in just a few minutes. For the Product Hunt community, we’re offering a 50% discount for the first month of our Team Plan, too. Just enter the discount code FABIPRODUCTHUNT when signing up for the Team tier. Finally, here are a few areas we'd love love your feedback on: - We’re considering adding an “AI agent” cell. This cell would take any number of inputs and a prompt and generate the type of output you specific. Think of this as an AI analyst agent that can run on a schedule with predictable inputs. Would this be interesting? - Are there any other sources or destination you would like to see us support? - How does the workflow canvas work for you? Any UX feedback welcome. Excited to hear what you all think!
Eric Wu

This product seems to enable better data visualization and simplify the data analysis process. Looking forward to tje product!

Marc Dupuis

@abandon_wu1 let us know what you think of it Eric! We love feedback

PriscillaIvy

Build, deploy, and share specialized AI data analyst agents no coding required. Automate insights, generate reports, and make data-driven decision faster then ever.

Enzo Avigo (📊 june.so)

Congrats on the launch!!

Marc Dupuis

@enzo_avigo thank you!

Hal Zeitlin

I used to have a really annoying to maintain retool db to do analysis and reports. I've been wondering when I can replace that. Now's the time.

Marc Dupuis

@hal_zeitlin2 100%. Retool is great for internal tooling, but there's definitely a better way for internal reports!

Ayush Agarwal

oh this is amazing. Just tested with a couple of queries and it works pretty well.

I have a couple of questions:

  • who is the ICP for this product?

  • How are you managing the data security? is it safe for organizations?

@mfdupuis @leitang

Marc Dupuis

Hey Ayush! Great questions:

  • ICP: We're used and loved by data practitioners at small and growth startups. So typically that's either the first data hire (a lead data analyst or data engineer), a PM or even the founder. We're also used by larger organizations that have an established team of data analysts and want to boost productivity and collaboration.

  • Security: We take this very seriously with a number of enterprise-grade measures, which I won't get into too much detail about here (feel free to reach out!). Always happy to share our policies, architecture and SOC2 Type 2 report with enterprise customers.

Saqib Butt

I love how Fabi.ai brings SQL and Python together in one platform — that combo alone makes me want to ditch juggling multiple tools. This could seriously speed up how I work with data.

Marc Dupuis

@saqib_butt love to hear it, that's the goal! Definitely let us know what you think when you get a chance to take it for a spin.

Amar Goel

looks really cool!

Sam Seely

Congrats on the launch, Fabi!

Marc Dupuis

Thank you Sam!