Case Study
Known for its rising reputation as a next-gen AI disrupter, Fetcherr specializes in leveraging data-fed generative AI models to streamline airline ticket pricing. The company was scaling at lightning speed within a native cloud ecosystem, but rising Google Cloud Platform (GCP) costs began to weigh them down. Sela delivered the cost control and flexible resources Fetcherr needed to scale with ease.
The Challenges
High costs and a lack of GCP knowledge
As Fetcherr’s vice president of research and development, Uri Schatzberg has always been on a mission to craft exceptional products, including top-tier generative AI technology. This tech can seamlessly process market data into large models that can facilitate reasonable, actionable decisions, especially about airline ticket pricing.
“We’re a cloud-native AI company,” Schatzberg said. “All of our pipelines, all of our value, everything is in the cloud. No local servers. That’s how we deliver speed and scalability.”
But relying solely on cloud infrastructure introduced issues, like a lack of GCP expertise.
“We use the full Google stack, but when we ran into specific technical issues, we had no one to ask,” Schatzberg said. “We’d have to hunt down knowledge, search forums, or bring in short-term contractors.” It could take months to exhaust these efforts and discover the root cause of an issue.
They also faced rapidly rising operational costs. Fetcherr’s cost base correlates directly to its cloud use expenditures, and those costs ultimately trickle down to customers. Schatzberg and his team needed a way to stay competitive without building a full-time FinOps department.
“It was becoming harder to understand what we were paying for and even harder to predict what we would pay next year.”
Instead of hiring more full-time employees, the best path forward involved finding a knowledgeable partner to reduce costs, enhance their GCP usage, and keep the company moving without breaking things.
The Solution
Sela offers FinOps, essential architecture, and Google experts
Some of Fetcherr’s affiliates recommended Sela to solve their scaling woes. Fetcherr thought Sela was the best option to address their immediate cost challenges, and it wasn’t long before they realized the team could offer much more. Suddenly, Fetcherr had a multi-dimensional super-resource at their fingertips that could evolve alongside their needs.
Instead of wasting previous developer hours troubleshooting cloud infrastructure issues, Fetcherr finally had a way to get the answers it needed quickly, via regular standing meetings and real-time collaborations.
“Right away, Sela gave us direct access to Google experts and their own deep bench of technical specialists,” Schatzberg said. “It was the first time we had someone to call when we got stuck.”
Fetcherr worked with Sela on a number of issues, such as saving compute power and optimizing Kubernetes. However, they prioritized their biggest cost: data storage. One of the partnership’s earliest wins involved mapping the company’s entire cloud expenditure. Sela’s experts helped the Fetcherr team pinpoint where they were “bleeding money” and needed to adjust.
“There are several payment strategies for different services, and Sela helped us define the right strategy for each service,” Schatzberg said. “That really made a huge difference in the expenditure.”
From there, Schatzberg and his team restructured their workflows to become more effective.
“They did everything. They showed us where the pain points were and how to fix them. They helped us budget and plan for the future. We didn’t need internal FinOps. Sela handled it all.”
Sela also helped guide Fetcherr through the detailed, complicated process of becoming a GCP partner — a major milestone for any company looking to build a business foundation using Google’s ecosystem.
“They were with us every step of the way,” Schatzberg said. “They even flew with us to meet Google executives. That level of dedication is rare.”
The Results
Superior cloud optimization that facilitates market-ready AI products
With Sela by their side, Fetcherr cut their cloud storage costs by around 50%. This reduction was a massive accomplishment, especially since they attribute roughly half of their cloud spend to storage.
Once cloud operating costs began to drop, Fetcherr’s product pricing model followed suit. The company could finally introduce more competitively priced solutions to customers without cutting corners on innovation or performance quality. The technical issues that used to take developers months to solve now take days.
“Sela helps developers do their job more easily,” Schatzberg said. “Instead of developers needing to understand what the problem is and trying to find the knowledge to fix it, Sela helps us understand how to solve the problem and connects us with the right experts.”
And because they solve problems quicker, they can develop solutions faster and deliver more value to customers.
Fetcherr’s relationship with Sela eventually grew beyond FinOps solutions and GCP management. Sela became the company’s go-to resource for system architecture support, emerging proof-of-concept projects, and ongoing technical advice.
“We’re even discussing bringing in a data architect through Sela. They’re so diverse — it’s like having a Swiss army knife in your back pocket.”
For companies using GCP and facing complicated challenges related to tech support and reliable cost control, Schatzberg recommends Sela.