Console

Interview with Evelyn LaTour

CPO, AutoCloud - Viualize cloud resources.

What is AutoCloud? Why did you build it?

AutoCloud is a cloud visibility tool. For enterprises that are running workloads on AWS, Azure and Google Cloud, we’re helping their engineering and DevOps teams maintain visibility into things like security, compliance, spend, and drift with three main features: automated technical documentation, interactive 3D visualizations, and a managed GraphQL API.

Our product is incredibly UX-focused and highly visual, which is uncommon in the world of developer tooling. We use auto-generated diagrams and visuals as the foundation upon which we layer insights and show connections and relationships between services. Our aim is to reduce the complexity and the time to insight by visualizing things that matter to the end-user in a human-readable way. Our grand vision with all of this is to reimagine cloud visibility by creating a new industry standard for visualizing and querying multi-cloud data.

AutoCloud has four founders Tyson, Drew, Chris, and myself. Tyson, Drew and Chris have worked together previously in a cloud consulting company and were helping enterprises with cloud migrations and greenfield projects where they kept encountering the same problem.

These companies didn’t have a core competency in technology and were really struggling to build reliable, secure, and cost-effective cloud infrastructure. They all had super small or completely non-existent devops teams. They couldn’t retain good talent and if they were hiring external consultants to come in and help out, things would be fine for a while. Once the consultants left, however, they had no idea how to maintain this infrastructure that had been built for them so it would degrade over time.

With AutoCloud we want to help these types of companies and create a tool that could give any developer with varying levels of knowledge about cloud infrastructure the abilities of a senior solutions architect. Even for the more experienced devops professionals, the aim is to make their jobs easier by giving them access to the data they want in the format they need.

How does it work alongside tools like Terraform?

We do quite a bit of integration with IaC. For some of our enterprise clients, we have this tool we internally call “Terra-Forms” where we take a pre-existing Terraform module, and create a front-end so that it’s really easy for them and all of their developers to deploy resources in a consistent manner that complies with their internal compliance policies. We love IaC and we’re looking for ways continually to incorporate it into the platform.

What does a “day in the life” look like for you?

One day is never like the next, as I’m sure is the case for most startups. First thing in the morning, I always work out otherwise the whole day is probably off to a bad start! I have amassed an enormous amount of workout equipment through Amazon Warehouse Deals over the pandemic, which has been wonderful. I then shower, have coffee, breakfast, and do a remote standup with all of our co-founders that we always have every morning. We’ve been a remote company since day one, so we’ve never had a real office. Fingers crossed maybe one day we’ll get a physical office.

Typically, the CEO and I will argue about some things since our desks are about 15 feet apart from each other, then I’ll check email, look at Jira, see what’s a super high priority and what I have to do during the day. I try to keep meetings to three hours or less, but sometimes that’s not feasible. There’s lots of planning involved in the CPO role, but I also have a background in user research and design, so I’m wearing a lot of different hats.

On any day, I could find myself doing product strategy, team management, front-end design mockups in Figma or creating the 3D models in Blender. I’ll also do code reviews, and then jump over and do social media strategy, handling our email lists, our email nurturing campaigns and content generation with PowerPoint decks, graphic design and copy editing.

I manage our landing pages for both AutoCloud and CloudGraph properties, including SEO analytics. I’m fully trained in user research and user outreach, doing interviews and translating those to findings, making sure it’s on our product roadmap. I still do a little bit of software development for the SaaS platform and 3D environment, but as time has gone on I’ve been trying to take a step back from all that.

As we work from home the day never really ends. I try to wrap things up around sixish at night, but I’m still available on Slack, looking at our danger channel for Sentry errors that look weird. I have no boundaries between work and life, but hopefully one day that’ll change.

What is the team structure around AutoCloud?

Right now we have 17 employees, which is great for a seed-stage company. We have seven people in the US, and everyone else is in South America. Really amazing engineering talent in Mexico, Chile, Argentina. The majority of the team speaks Spanish!

The team structure is based on the products themselves. We have the AutoCloud team for the SaaS platform mainly consisting of full-stack engineers, and within that, we have a mini 3D environment team, which is just me and Paul Elliot, who’s a fantastic 3D web dev.

Next we have our CloudGraph team, which is our open source team. Under the hood, CloudGraph is powering AutoCloud. The backend of AutoCloud gets all of the data that we need from the cloud provider SDKs and then sends it over to AutoCloud. AutoCloud is formatting it for the diagram layouts, the compliance and security scans, the drift comparisons and things like that. They share the same GraphQL API, but AutoCloud has a more managed version. Each of those teams has its own separate sprint planning and scrums ceremonies.

Lastly, we have our infrastructure team who are responsible for keeping all of our systems up and running, maintaining our SOC compliance, CI/CD pipelines, etc.

We also have one superwoman of a project manager Anabella who just keeps everything running across all of the teams. I don’t know how she does it, but before her, our life was a complete mess!

How did you first get into software development?

I actually have a master’s degree in Cultural Anthropology. In college I really wanted to go into software development because I was into playing lots of video games. I went to school in Canada, in British Columbia, that had a very old school approach to software development. You had to pass a theoretical mathematics course in your first year to even be allowed into the software development program. I failed that course harder than anything I’ve ever failed in my entire life.

To this day, now having been a software developer for coming up on four years, I still have not seen or come across anything in that theoretical mathematics course. I don’t know why that was a requirement, so that destroyed my early dreams of being a developer. A decade later I find myself now starting a software company.

It was a very long route to get there. I got my master’s degree, then I started working in market research for a pharmaceutical company. We started getting a ton of projects that would come in where these pharma companies were building their own apps for mobile devices, for physicians and for patients. They also had websites that they had no idea how to design or do anything with. I was asked to figure out what to do and how to have an effective digital communication strategy. I took on all those projects, and taught myself from scratch, learning how to do UX research and UX design, and found that I liked it a lot.

I then moved over to Motorola and was leading user research there for the cameras on the phones. That was super fun, but as everyone knows, Motorola Mobility was not doing super well. While at Motorola, I had the chance to do a software development boot camp at Fullstack Academy in Chicago.

I learned everything I needed to get started and then went to work for a prop trading firm, but found that I wanted something different working there. Having come from a role where I was interfacing with so many people, I love software development, but I also don’t want to be taking tickets and just knocking them out.

I had a side project working with Tensorflow.js and leveraging Google’s system that they put together called MobileNet that would classify images, so you could basically write on top of that and create your own image classification. I had always wanted to start my own company and Tyson (the CEO) and I had been together for a few years at that point. I was planning an app for the retail fashion industry, but then we decided, we’re probably better just combining forces and working on a project together. AutoCloud was born.

What is the most interesting development challenge you’ve faced working on AutoCloud?

The original algorithm for laying out the infrastructure diagrams and 3D space was probably the most challenging. Our original idea was to create these Terraform generation templates, we realized after a few months this was going to require a huge team of people. We’d been doing a lot of interviews with our customers, and one of the things that had come through was could you just show me everything that I have?

We knew how to get all the data but we didn’t really have a way to visually lay it out. Coming from design, UX, and then also software development, I felt this was an interesting problem to tackle. We did it in 3D to take it to the next level. I had to learn the 3D environment API and we leveraged Amazon Sumerian under the hood for the 3D environment. Sumerian has its own special way of creating meshes, handling asynchronous and synchronous actions, like how do I create this mesh, and then I have to move it and I have to copy it.

We had to take a JSON object and somehow generate a visual representation of it. How do the entities know how big they should be? How do they know where they should live in relation to other things? That was a creative and fun problem to solve and I learned a lot about game design since all of the browser-based game engines operate on the exact same principles, which was cool. We could take the knowledge from say, Sumerian to ThreeJS, to whatever 3D web dev platform you want to work with and all transfer it across.

That code is still powering a lot of our visualizations, but of course, Paul has come and improved upon it.

What is the most interesting technology product or project that you’ve been playing around with recently?

I am continually impressed by Webflow. It is a tool that I was initially very opposed to using because I made our original landing pages from scratch with Gatsby, which is this really cool, full-stack framework for creating really powerful web pages. It handles a lot of stuff like prefetching and lazy loading and image optimization. It handles all this stuff for you. It’s a really great tool.

However, you’re still writing React components and spending a lot of time on something that maybe isn’t the core of your business, especially since for us, should not be spending all my time writing a landing page from scratch. Most it was just our landing page. You have to think, we’ve got a couple of people, I startups shouldn’t be doing that unless they’re using Gatsby for their actual product. I was pushed over to Webflow and it is so much better than using something like WordPress or Squarespace.

The amount of customization it allows and the amount that they’ve stayed true to just basic HTML concepts is ideal. They still have margin and padding and all those things, and they have CSS Grid and Flexbox. All of these concepts that as a front end developer, you are completely accustomed to working with, you can do it in Webflow and you can get rid of all of the hassles of having to have a build pipeline and involve your infrastructure team in making sure that the website is continually up, deployed when you need it. They don’t have to deal with it at all.

They set up the hosting and then even someone that’s non-technical can go in there and edit some text. They even have a CMS. It’s really been a great way for us to build an awesome landing page that’s super customized.

What is the rest of the tech stack?

We use React on the front end and Amazon Cognito for all of the user signups and all of our handling of that data. I believe we’re still using Express and Node.js in the backend, and then we have GraphQL and Prisma for API interactions.

Describe your computer hardware setup

I’ve been using the same 15-inch MacBook Pro since 2017. It has been a workhorse, but as of late, there have been some difficulties with it keeping up with all of the memory-intensive applications. Try running Adobe Illustrator, Figma, a 3D environment and a Zoom call at the same time. The computer just shuts down.. I think I’ll upgrade this soon.

In terms of monitors, I’ve got a 34'' Samsung Ultrawide QHD, so that I could have enough space horizontally to have three files open on VS Code at the same time.

Next to this monitor, I have my favorite child, hopefully Samsung won’t be listening, but it’s my LG 4K IPS Monitor, which is gorgeous and beautiful for doing all of any kind of design work. The color depth on the IPS panels is amazing. Eventually, I’d like to have two of those, having the VA panel and the IPS side by side is not recommended since their colors are completely different.

Mouse-wise, I have the Logitech MX Master 3, which is perfect for development or design work, and it has the perfect FreeWheel scrolling. You can go through thousands of lines of code in one swipe and stop on a dime. It has horizontal scrolling and everything is customizable and is mapped to macOS, so I can just go to mission control and switch between screens really quickly. I have endless fights between myself and the CEO who still uses the Magic Mouse from Apple without a right-click button. Although I’m still using the Apple keyboard with the number pad on it, hopefully, we’ll upgrade that at some point, because the ergonomics are not the best. Headphones wise, I use BOSE QC wireless active noise canceling earbuds, which are nice.

Describe your computer software setup

OS: macOS

Browser: Chrome. Although I also use Firefox, Edge and Safari for all the QA testing on the platform

Email: Gmail.

Chat: Slack.

IDE: VS Code.

Source control: Git and GitLab.

Describe your desk setup

I’m very particular about my chairs because I like to sit cross-legged or put my leg over an arm. So I had to find one that had flared arms that were just part of the chair and also cushioned. I found one on Amazon and I haven’t looked back ever since. Other than that, I don’t have anything special.

The desk of Evelyn LaTour, AutoCloud

When coding

Daytime or nighttime? Daytime.

Tea or coffee? Coffee.

Silence or music? Silence or podcasts, depending upon how much thinking the work that I’m doing requires.

What non-tech activities do you like to do?

Working out and yoga. I do a lot of cooking, which has been easy especially during the pandemic when you can’t really do anything else.

I’ve always been into investing, trading and my friend recently got me into crypto and DeFis, which has been fun for me as well. I love to travel, and Tyson has gotten me into skiing and scuba diving. I’d love to get certified one day.

Find out more

AutoCloud allows you to visualize your cloud resources. It was featured as an “Interesting Tool” in the Console newsletter on 13 Jan 2022. This interview was conducted on 10 Jan 2022.