Speedscale: How This Company Helps Engineers Build Solid Kubernetes Apps

By Amit Chowdhry • Mar 16, 2023
  • Speedscale is a company that helps engineers build Kubernetes apps. This is the story behind Speedscale.

Speedscale is a Kubernetes traffic replay company that simulates production conditions that recently announced it raised $9 million in a combination of pre-seed and seed funding. Grotech Ventures led the investment round with participation from Sierra Ventures, Tech Square Ventures, Correlation Ventures, CreativeCo, Tom Noonan, Wally Wang, and TIE Atlanta. To learn more about the company, Pulse 2.0 interviewed Speedscale CEO Ken Ahrens (pictured below).

Ken Ahrens

How The Company Came Together

When I asked Ahrens how the idea for the company came together, he said that the team comes from a long background in observability and monitoring at companies like New Relic, Observe, and CA Technologies.

“We noticed how many retail, telco, fintech, media, and SaaS companies waited until customer experiences were degrading before they remediated – typically with a rollback or a large troubleshooting session with all hands on deck.  But why was that? Software is difficult to validate, given the number of microservices we’re adding and the pace at which we push updates.  It’s tough for engineering leaders to have confidence in the changes they’re pushing. Coupled with the lack of innovation in the software testing space, we wanted to revolutionize the way companies validate performance and functionality,” said Ahrens.

The team put a few ideas on a whiteboard and then had a healthy discussion about the shape and ergonomics of the solution. And a software executive once told them that they started the Agile and DevOps journey “but forgot to bring software quality with us.” And they acknowledged a desire to help engineering teams build resilient apps and allow them to do so quickly.

Following a couple of initial prototypes and proofs-of-concept, they were actively pursued by Y-Combinator to enroll and then they learned a lot about the ideal user persona, GTM, and acquired their first customer shortly after.

Biggest Milestones

I asked Ahrens what some of the company’s biggest milestones have been so far. And Ahrens pointed out that as a startup, they are able to move quickly and the company went through a couple of major iterations of their product so far.

“Every time, we unlock new architectural scale and acquire new customers. When we started in January of 2020 and the pandemic hit in March, we were worried like the rest of the world. But little did we know, touchless/curbside delivery and digital commerce were about to accelerate in a tremendous way. We were quickly able to build to $300K annual recurring revenue (ARR) with a couple of big name customers and started to listen to feedback regarding their favorite features.  Continuing to work with our users, and removing the rough edges have been paramount to our success. We’re very blessed to have the opportunity to help large enterprises such as Sephora, Vistaprint, and Zenni Optical (among others), which allowed us to raise $9 million in seed funding at the end of 2022,” added Ahrens.

Market Solutions

What are some of the market challenges that Speedscale solves? Ahrens pointed out that engineering and development teams often spend hours or days manually testing their APIs and apps, which has been a longstanding challenge. And despite strides in DevOps and monitoring, organizations still end up in the news due to production incidents and outages.

“We built our company and product on the premise that successful cloud-native application delivery can only happen in predictable software-defined environments. Speedscale provides an entirely new automated approach that’s changing how developers validate new code and flag and resolve problems ahead of a release. For the first time in the space, we’re providing a modern way to scale API quality automation as fast as releases take place, combining observability technology with cloud data warehouses. We’ve been fortunate to gain rapid traction as an enterprise industry player within the Production Traffic Replication (PTR) space, exclusively helping Kubernetes engineers stress-test their APIs with real-world scenarios,” explained Ahrens.

Core Products And Features

In capturing production traffic, sanitizing any Personally Identifiable Information (PII) and replaying it, Speedscale has 3 core features:

1.) Traffic Viewer – To browse recorded traffic and “pick” the transactions to recreate the desired conditions such as business as usual, Black Friday, lunch hour rush, benefits enrollment day, a new stock IPO, etc..

2.) Traffic Generator – This feature allows you to exercise your app in a variety of ways like soak testing, contract/regression validation, high load simulation, or even chaos

3.) Mock Server – Everyone integrates with third parties like UPS, Google Maps, Stripe, Twilio, etc. But the developer environments may not be as realistic or performant as the real thing. And a simulator autogenerated and modeled after real traffic conversations can serve as a stand-in for constrained or otherwise unavailable systems.

Pursuing Kubernetes

When I asked Ahrens about why they decided to pursue the Kubernetes platform, they said that resiliency and scalability are as important as ever.

“In a 2021 Cloud Native Computing Foundation (CNCF) survey, 96% of respondents reported evaluating or using Kubernetes, with 3.9 million users worldwide (67% more than the previous year). However, Kubernetes applications are growing more distributed and complex and, as a result, conventional validation techniques can’t keep up,” stated Ahrens. “PTR allows developers to do much more without scripting, reduce cloud costs, and ensure application performance. Engineering leaders have long awaited a way to simulate realistic conditions, capture actual sanitized data scenarios, and stand-up performance environments that behave like they do in production. Lastly, Kubernetes is a great platform with built-in network rules, autoscaling, and patterns for workload management that allow Speedscale to maximize the ephemeral deployment of traffic replay pods.”

Evolution Of Speedscale’s Tech

When Speedscale first started developing its technology, the original thesis was that enterprises had adequate ways of exercising their applications but lacked the environments in which to do that in. So the team started out focusing on mocking capabilities, which simulate conversations with downstream dependencies.  But then they quickly found that generating ways to invoke applications and validating functionality was manual, time-consuming, and required building large swaths of scripts. Those scripts then needed maintenance and modification every time you released them. Now The Production Traffic Replication service covers both exercising and simulating environments to expedite all aspects of gaining confidence in your changes.

Plus the ability to drill down into transactions in the Traffic Viewer and study the message payloads has been invaluable to customers in root-cause analysis of problems. And while monitoring tools like Grafana, DataDog, and New Relic can aggregate metrics and metadata surrounding such transactions, Ahrens pointed out that they learned API Observability is a much-needed functionality that may not be available in some developer environments.

Who Are Speedscale’s Customers?

Speedscale’s customers include a growing number of SaaS providers and Fortune 2000 companies. And these customers include recognizable e-commerce and CPG brands like Sephora, Vistaprint, and Zenni Optical. More specifically, using the Traffic Replay solution, Vistaprint reduced load testing times by 80%. Vistaprint is a global e-commerce company that produces custom marketing materials for millions of customers worldwide and they experience high volumes of traffic for customers looking to design and customize prints on business cards, signage, and promotional items.

“As mentioned, the best way to make sure new code is validated is to ensure that it runs as code is checked in. Another large Speedscale customer, Sephora, is constantly innovating around their shopping experience.  Their CTO Sree Sreedhararaj states ‘With the exponential increase of microservices and capability-focused development, a platform like Speedscale is the only way to certify and confidently deploy these features to our customers at scale and speed,'” Ahrens said.

What Makes Speedscale Different From The Competition?

Ahrens pointed out that Speedscale is founded on the unique insight that software validation is impossible without 3 essential elements – the proper transactions, data, and environment. And most organizations focus on one element at a time: 1.) Postman for transactions, 2.) some sort of Test Data Management solution, and then 3.) large end-to-end environments in the cloud.

“But by orchestrating all 3 elements in unison, you can dramatically accelerate software delivery.  Moreover, by autogenerating the transactions and data from real, sanitized traffic, you eliminate a lot of the toil associated with conventional test scripting,” Ahrens commented.

With Speedscale:

– Engineers can exercise their app by orchestrating tests, data, and mocks for a complete validation service.

– Engineers can leverage traffic to autogenerate necessary artifacts and model behavior based on real data sets.

– Engineers have a self-service platform through which they can understand API dependencies and inspect message payloads

– Engineers can run a variety of validations, such as contract, regression, load, and chaos, with one tool.

Speedscale’s Future Goals

The first job will be to raise market awareness that Traffic Replay is a standard part of the CI/CD (Continuous Integration, Continuous Deployment) process. And exercising an application with regular Traffic Replay tests dramatically improves time to market, quality, and efficiency. The idea of Traffic Replay is intuitive but many decision-makers are not aware that it is technically possible because until recently, it wasn’t. Key enabling technologies such as Kubernetes, inexpensive cloud storage, and AI have advanced.

“We’ve seen that as our customers move to Kubernetes they don’t tend to improve one tool or process in isolation. Instead, they choose an entirely new ecosystem. This fundamental shift is an enormous opportunity for Speedscale and its partners. Speedscale’s technical approach dramatically improves software validation, but the core technology also enables improvements in securing and monitoring applications as well,” Ahrens revealed.