Scaling to $7.5bn Exposed the Fault Lines in the Modern Data Stack

Editor’s note: Today’s conversation features Mitch Bregman, founder of Galaxy, a company taking aim at one of the most fragmented and expensive parts of the modern enterprise: the data stack.

Galaxy is building what Mitch calls a “data platform in a box,” a single pane of glass designed to replace the 6–12 tool stacks many companies stitch together at a cost of $1mm–$10mm+ per year. Mitch and his co-founder, Leon Kozlowski, helped build this type of system as early employees at Flock Safety, a $7.5bn company focused on public safety.The thesis is bold: most of that fragmentation is unnecessary, self-serve analytics still is not solved and AI plus open source finally make consolidation possible.

There are strong lessons here for buyers and operators around stack sprawl, usage-based pricing, human-in-the-loop AI design and what happens when the line between technical and non-technical roles starts to disappear.

As always, let us know what you think and please reach out if you have feedback.

- Dave

Enterprise teams know the problem: data lives everywhere. Ingestion tools, warehouses, orchestration layers, BI dashboards, reverse ETL and point analytics products. Every new question requires another handoff. Every new tool adds another contract. And the promise of “self-serve” usually ends in a ticket queue.

That is where Mitch Bregman [LinkedIn] and Galaxy come in. Galaxy aims to collapse the modern data stack into a single observable system, automatically building a semantic layer on connect and enabling both technical and non-technical users to ask questions, build workflows and operate from the same pane of glass. As their tagline states, Galaxy helps to unify definitions, processes and meaning into one shared model that AI can reason over and people can trust.

In this conversation, Mitch covers:

• Why the current data tooling ecosystem is “largely a scam”
• What 6–12 point solutions actually cost mid-market and enterprise companies
• Why Galaxy is launching web-first with a product-led motion
• How AI agents plus human approval beat free-roaming automation
• Why usage-based pricing makes more sense than seats in AI infrastructure
• His spicy 20-year take: point solutions disappear and software builds itself in real time

If you care about reducing stack sprawl, aligning business and engineering teams or preparing for a world where “business engineers” replace rigid role boundaries, this interview explores what it really takes to simplify data without sacrificing control.

Introduction

Dave Ambrose

Mitch, thank you so much for doing this. I’m really pumped that you’re here. I hear this is your first ever kind of podcast-style, kind of interview-style format. We’ll take it easy today. It’ll be pretty low key and chill.

So Mitch, I think I know a decent amount about your background and the problem that you’re solving, but why don’t you give the audience an overview of what Galaxy is - meaning the product that you’re building - and then also the problem that you’re solving for Galaxy.

What Galaxy is: “data platform in a box” and why the stack is broken (00:49.166)

Mitch Bregman (00:49.166)

First off, thanks for having me, Dave. I’m super excited to be here. So yeah, my name is Mitch Bregman. I’m the founder of Galaxy.

Galaxy is aiming to build a data platform in a box. Effectively, building a data platform is very, very hard. You procure a lot of different tools to build a data platform. Think about the sources, the warehouses, the analytics that must happen on those warehouses, the ingestion layer, the transformation layer. And our view is that this fragmentation in the data space is actually largely a scam.

And you have a lot of niche players who are building these various different sorts of niche products in these different spaces to kind of create this engine - which is a data platform - that effectively does something pretty simple. It’s just asking a question and getting an answer, right? At the end of the day, business folks need to get answers to questions to make decisions, especially correct decisions.

Right now, the cycle to get an answer to a question is very vicious. They go to an analytics engineer, and an analytics engineer goes to a data engineer, a data engineer goes to a software engineer. We want to create a single - our goal is to create a single pane of glass - where every single person on your team that needs data or needs to interact with data is able to do it in a very streamlined, flexible, and easy way to get the answers that they’re looking for, to build the apps on top of the data that they have, and create workflows that are reproducible, observable, and allow them to create efficiencies in their day-to-day work.

Why now, and how do you sell it? (02:14.152)

Dave Ambrose (02:14.152)

So, I mean, what you just described is the dream for every CEO, CFO, procurement director. And I’m curious, because I think this has been tried before - which you and I have kind of discussed in different informed factors.