A double unicorn transforming supply chain visibility

Valued at $2.7 billion, project44 is an advanced visibility platform dedicated to supply chain companies, from shippers to third-party logistics firms. 

The cloud-based platform connects, automates, and provides visibility into crucial transportation processes to help companies increase operational efficiency and deliver an outstanding, Amazon-like experience to their customers.

project44 continues to innovate and deliver new services to its customers, registering 70% year-over-year growth. So far, the company has raised $897.5 million in multiple funding rounds from global players in logistics solutions and investors such as Goldman Sachs Asset Management, TPG, Emergence Capital, Chicago Ventures, Sapphire, 8VC, and others.

To accomplish its business goals, project44 needs a cloud infrastructure that meets strict requirements.

Since we’re a truly global company both in terms of our footprint and customers, our infrastructure has a lot of requirements around processing massive volumes of data. Other key aspects are availability, performance, and security, as the processes we support are business-critical for our customers.

Our footprint reflects that. We’re multi-cloud and multi region, we have deployments in a great number of locations. And we have very high SLAs in terms of time, reliability, and performance. I’d say that we have a pretty complex and business-critical infrastructure that represents a significant portion of our spend.

Olivier Modica, VP of Engineering at project44

The primary objectives of the infrastructure team are to provide a reliable infrastructure for product teams to build products and to use modern technologies to achieve this. “We want to be able to provide current capabilities. Around 50% of our infrastructure runs on Kubernetes,” said Olivier Modica.

A push for cloud cost optimization

Similar to other technology companies, project44 set a goal for its teams to make the best use of its capital and resources. “There’s definitely been an increased focus on cost visibility and cost optimization across the entire organization,” said Olivier Modica.

We are a high-growth company and need to be careful about prioritizing our work. One area we wanted to focus on was getting better and more granular data about our infrastructure costs. We’ve been struggling with getting a bottom-up, granular view of how Kubernetes resources are being used and whether we are reporting them accurately.

Olivier Modica, VP of Engineering at project44

Initially, the team focused on cost visualization and resource utilization. But soon, cost optimization became a priority as well. This motivated project44 to start looking for a solution that would meet all of these objectives.

Obviously, we could achieve some of that using native tooling from cloud providers. But we knew that it would take massive effort to make it work internally. So, we started looking for something better. We did a bake-off and assessed a few vendors, considering security, reliability, and speed of achieving savings results.

Olivier Modica, VP of Engineering at project44

project44’s VP of Engineering, Olivier Modica, and his team met the CAST AI team for a platform demo.

We were impressed with the expertise and research done by the CAST AI team. We have some technical experts on our team who are capable of realistically judging different solutions. I think everyone was quite impressed with CAST AI.

We were looking to achieve results in weeks and months, not years. Once we completed the POC with CAST AI, we could confirm that this was the right vendor for us. Getting a new budget was easy since we had a very strong argument based on cost reduction.

Olivier Modica, VP of Engineering at project44

Achieving 50% of cost savings within one month

Speed in delivering results was one of the key requirements of project44 and CAST AI matched it right from the start. “When assessing vendors, we’d ask ourselves: is this a solution that we could deploy fairly effectively and quickly? Or would it still require quite a heavy lift from our team? CAST AI was definitely the right choice,” said Olivier Modica.

It took us a month from meeting the CAST AI team to closing the POC. We’ve rolled out the solution gradually over three or four months, initially on lower environments and not production clusters. Everything was working as expected, which gave us the confidence to move forward. We achieved savings practically immediately, so our ROI timeline was very short.

Olivier Modica, VP of Engineering at project44

By implementing CAST AI, project44 saw 50% of cost reduction on compute costs within the initial rollout cluster during the first month. Next, the team onboarded five more clusters and achieved 46% of cost savings within the initial phase. This is just the beginning since the company has a roadmap for achieving further cost visibility and savings opportunities within the second rollout phase.

At every step, project44 could choose from a spectrum of available savings with the CAST AI platform based on the configuration choices. “We are definitely well within what we expected right now. There have been no surprises along the way, which is a great thing. And we’re confident that we can increase our savings by being a little more aggressive in cost optimization – for example, by using more spot instances,” said Olivier Modica.

The key value of CAST AI is that it’s a very easy sell. And the ROI is pretty much immediate. The platform is very beneficial because it achieves savings automatically and because of the way the pricing is set up. I think the solution sells itself well – it makes everything easier and enables teams like us to move very quickly.

Olivier Modica, VP of Engineering at project44

CAST AI fits right into project44’s FinOps initiative

What made a real difference for project44 was the Cost visibility at the level of granularity adjusted for the unique needs of teams working with Kubernetes.

The most exciting part of the CAST AI platform from a business case standpoint is the cost savings, of course. But what’s been very valuable and useful for us is a granular view of how we’re using our resources. 

Olivier Modica, VP of Engineering at project44

The ability to slice and dice the workloads allowed the team to learn how much they could do from an optimization perspective. The bottom-up visibility lets everyone involved check the number of CPUs and GBs a given cluster, workload, or namespace is using.

We can check if the resources provisioned for the cluster are of the right size and use that information to enable FinOps capabilities across the business. It also gives us a good foundation for having a conversation with product teams and providing them with additional insights about infrastructure performance.

Olivier Modica, VP of Engineering at project44

A strategic partnership with CAST AI

The CAST AI team supported project44 at every step of the implementation. “The support has been fantastic. We have a direct line with the team, and the process has been very efficient. CAST AI has a lot of Kubernetes experts on board, and we had the opportunity to have some discussions about how things work or what’s the best choice,” said Olivier Modica.

Providing my feedback to engineering teams has been great as well. We’ve been communicating with CAST AI about the features we’re using and some potential improvements. The team has been very receptive to feedback, so I feel it’s a very collaborative relationship, which is fantastic.

Olivier Modica, VP of Engineering at project44

CAST AI has become a key vendor for project44, delivering cost savings and Kubernetes automation capabilities that strengthen the company’s position as an industry leader.

We built a truly collaborative relationship between our two companies. The quality of the team and its responsiveness have been very important to me because this is the kind of relationship we want to have with strategic vendors.

Olivier Modica, VP of Engineering at project44