By Cloud development cycles in regulated environments routinely consume 30 minutes or more per deployment. Developers in banking and insurance often have no direct cloud access at all, forcing every change through a CI/CD pipeline before they see a single result. As AI agents begin generating production code at volume, the cost of slow or unsafe feedback loops compounds rapidly.
In this interview on TFiR, Waldemar Hummer, Co-founder and CTO at LocalStack, walks through how LocalStack provides a fully local cloud development environment that compresses feedback cycles from hours to seconds, and why that same infrastructure now functions as a critical sandbox for agentic AI workflows.
Guest: Waldemar Hummer, Co-founder and CTO at LocalStack
Show: TFiR
Here is what every platform engineer and cloud developer needs to know.
Technical Deep Dive
Q: What is LocalStack and what problem does it solve for cloud developers?
Waldemar Hummer, Co-founder and CTO at LocalStack, describes LocalStack as a local cloud development environment that gives developers the ability to build and test applications entirely on their own machine. Rather than pushing changes to the real cloud and waiting through multi-minute or multi-hour deployment cycles, developers get feedback in seconds. The project began as open source in 2017 and became a funded company in 2021, incorporated in Switzerland with a globally distributed team.
“We provide this local cloud development environment. It gives you the ability to test your applications and build them fully on your local machine, on your laptop, giving you much quicker feedback cycles.” — Waldemar Hummer, Co-founder and CTO, LocalStack
Q: Why is the cloud developer experience described as subpar even when the cloud itself is reliable?
Hummer draws a distinction between the cloud as an infrastructure platform and the cloud as a development environment. The cloud is highly scalable, cost effective, and secure, but the developer experience of iterating against it is slow and friction-heavy. Every change requires a deployment cycle, which breaks the rapid feedback loop that good software development depends on. LocalStack addresses the experience layer without replacing the cloud as the production target.
“The developer experience is often subpar, not ideal and not optimal. Instead of having to push every change to the cloud and wait for several minutes to deploy, we give you the ability to have a much more snappy experience on the local machine.” — Waldemar Hummer, Co-founder and CTO, LocalStack
Q: How does LocalStack actually work for a developer making code changes locally?
Hummer describes a workflow where a developer edits code in their IDE, whether that is a Lambda function or a containerized application, and the change reflects locally in seconds without triggering any external deployment process. The feedback cycle drops from minutes to seconds. This removes the cyclic deploy-wait-inspect loop that normally separates a code change from a visible result.
“The moment they make a change in that application, it automatically reflects locally and they don’t need to go through any cyclic deployment.” — Waldemar Hummer, Co-founder and CTO, LocalStack
Q: Why do regulated industries like banking and insurance face a more severe version of this problem?
In regulated industries such as banking and insurance, developers frequently have no direct access to cloud environments at all. Every change must pass through a CI/CD pipeline before it can be observed, which Hummer says can take around 30 minutes per cycle. This makes the feedback loop problem significantly worse than in standard cloud development contexts. LocalStack allows these developers to test locally without requiring any cloud access, removing the pipeline as a bottleneck during development.
“Sometimes developers don’t even have access to the cloud. They might need to push changes through some CI/CD pipeline that takes half an hour to deploy, so the feedback cycles become even slower and even more tedious.” — Waldemar Hummer, Co-founder and CTO, LocalStack
Q: How does the rise of agentic AI development make the local cloud testing problem more urgent?
Hummer explains that as AI agents take on more of the code generation workload, the volume and velocity of untested code entering pipelines increases substantially. The problem is not just speed. It is trust. Agent-generated code cannot be assumed to be correct and must be validated before it touches a live cloud environment. A local sandbox that runs fast and is isolated from production becomes a necessary control layer in any agentic development workflow.
“You cannot always trust the code that’s being generated by these agents. You need a way to test this in an efficient and also sandbox manner without risking agents pushing changes to the real cloud environment.” — Waldemar Hummer, Co-founder and CTO, LocalStack
Q: What is the LocalStack AI sandbox use case and how does it work?
Hummer positions LocalStack as an agentic cloud AI sandbox, a local environment where AI-generated code can be executed and validated without connecting to or modifying any real cloud resources. The sandbox runs on a local machine and isolates agent activity from production systems. This use case extends LocalStack’s original developer productivity value into the AI workflow layer, where the need for rapid, safe validation is even higher than in human-driven development.
“LocalStack as an agentic cloud AI sandbox that runs on local machines.” — Waldemar Hummer, Co-founder and CTO, LocalStack
Q: How did LocalStack originate and how has it grown into a commercial product?
LocalStack began as an open source project in 2017 with the release of its first public version. Over several years the project grew in adoption, and in 2021 the team secured seed funding and formally incorporated the company. LocalStack is incorporated in Switzerland, backed by US investors, and operates with a globally distributed team. The commercial product builds on the open source foundation to serve enterprise development and platform engineering teams.
“It started in 2017 when we put out the first version of LocalStack and that has kind of grown over the years. In 2021 we got our seed funding and built the company.” — Waldemar Hummer, Co-founder and CTO, LocalStack
Resources & Documentation
- LocalStack, local cloud development environment for building and testing AWS applications on your machine
- LocalStack on GitHub, open source repository for the LocalStack project, active since 2017
***
👇 Click to Read Full Raw Transcript
Swapnil Bhartiya: Talk about the company. What do you do, what problem you’re trying to solve for the whole ecosystem, where you’re founded, when you’re founded. So it’s a history and a story.
Waldemar Hummer: It’s a pleasure to be here. So I’m CTO and co founder of LocalStack. So we’ve been around as an, as an open source project for quite a long time. Actually. It started in 2017 when we put out the first version of Local Stack and that has kind of grown over the years. And in 2021 we got our seed funding and built the company. We’re incorporated in Switzerland, but have US investors on board and our team is globally distributed. So what we do as a product is we provide this local cloud development environment. It gives you the ability to test your applications and build them fully on your local machine, on your laptop, giving you much quicker feedback cycles and reducing some of the friction and burden that you usually have when testing against the real cloud. So that’s our main value proposition.
Swapnil Bhartiya: When we look at Kubecon or the whole Kubernetes system now, the things are slowly moving towards agent AI, but this is still a kind of exciting space. And one of the things associated with Kubernetes is complexity, you know, and I mean we have kind of. It’s not that this complex is going to go away. I mean, now it’s maturist production. So we all know this is not going to go away. We have to learn how to deal with it. So can you talk about how does Local Stack make cloud software development easier and faster and more efficient for developers? So they don’t look at it like, oh no, let’s do something exciting.
Waldemar Hummer: The basic premise is that when you develop against the cloud, first of all, it’s this highly scalable system that has all security baked in and it’s cost efficient, cost effective, but the developer experience is often subpar, not ideal and not optimal. So what we do with Local Stack is instead of having to push every change to the cloud and wait for several minutes to deploy, we give you the ability to have a much more snappy experience on the local machine. So you’re talking not about minutes or even hours sometimes to deploy, but it’s more like seconds that the feedback cycle. So you can think of it as your developer maybe has some IDE in front of them and they make some changes to their code, be it a lambda function or some kind of containerized application. And the moment they make a change in that application, it automatically reflects locally and they don’t need to go through any cyclic deployment. This problem is even more aggravated by some regulated industries where let’s say you’re in a bank or an insurance company where sometimes developers don’t even have access to the cloud. So they might need to push changes through some CI CD pipeline that takes, let’s say half an hour to deploy so the feedback cycles become even slower and even more tedious. And that’s exactly what we are addressing with the solution. Now when you brought up agentic, the whole world of software is changing. We see a lot more agents these days doing the development. It’s very powerful tools. Of course. At the same time this even increases the necessity to test and validate code very quickly because you cannot always trust the code that’s being generated by these agents. You need a way to test this in an efficient and also sandbox manner without risking agent pushing changes to the real cloud environment. That’s where our AI sandbox use case comes in. So Local Stack as an agentic cloud AI sandbox that runs many local machines.
