Challenge
Kelvinlabs was trying to run a product-led growth motion on top of Clearpath's rigid data model, which buried usage signals and left the team guessing about which accounts were worth a conversation.
Solution
Make Perigee the system of record for product and customer data, then build the high-leverage workflows that route the right account to the right person at the right moment.
Results
Kelvinlabs' story
In 2021, Tomas Ferreira started building a way to run heavy compute jobs on demand — no cluster to babysit, no container images to hand-tune, no infrastructure to keep warm between runs.
The goal was narrow and stubborn: let a researcher iterate on an idea in the time it takes to make coffee, instead of the time it takes to learn an orchestration framework.
What started as a convenience for one team turned into a product. Kelvinlabs now runs compute-heavy, latency-sensitive work for several thousand developers, including teams at Northgate Pay, Junebug and Larkfield, and closed a €22.4M Series A in early 2024.
“Kelvinlabs is an infrastructure product, which is what makes it genuinely fun to sell to developers — you launch arbitrary cloud workloads, so the people evaluating us are extremely technical,” says Priya Raghunathan, Head of Revenue Operations at Kelvinlabs. “We work closely with teams building AI products and features, and we need them to find the right shape of the product fast and then quietly scale into it.”
Where Clearpath fell short
Every product-led company wants the same thing: one place that holds product and customer data together, so the team can see the signal in the week it happens, qualify accounts, build a durable picture of the customer, and cut down churn.
Like a lot of early teams, Kelvinlabs started on Clearpath — and then spent a year working around it.
“It was genuinely hard to set Clearpath up to reflect our data structures,” says Priya. “We model everything around workspaces, and workspaces are many-to-many with people. We ended up with a CRM full of floating records, duplicate companies, and reports that quietly disagreed with each other.”
Clearpath's inflexibility also created problems for Kelvinlabs. The idea of workspaces is really how the team tracks product data, and Clearpath simply doesn't have native support for that type of data. Nobody could ever be quite confident that what they read in the CRM was true.
Given the cost of Clearpath and the hassle of bending product data to fit its structures, Kelvinlabs knew they needed a CRM that could hold their model as it actually was.
The Perigee moment
When Priya first heard about Perigee, the reaction was immediate.
“When I opened Perigee for the first time, I had the feeling that this was the start of the next generation of CRM,” says Priya.
She was thoroughly unimpressed by how slow migrations usually are, so the pace mattered: “The fact that I could configure our CRM to fit how we work, rather than the other way round, in an afternoon — that was the surprise.”
Priya also loved the product design. “It's as quick and considered as the tools our engineers choose for themselves. I genuinely feel happy when I'm in it, which is almost unheard of in this category.”
How Kelvinlabs uses Perigee
Kelvinlabs uses Perigee as the central engine for their PLG motion. “Perigee gives us confidence that we have a foundation of customer data we can trust, which is what lets us build high-leverage workflows on top of it,” says Priya.
A single source of truth for customer and product data
The native Workspaces object has let Priya shape Perigee around the Kelvinlabs data model instead of the reverse.
“Our whole product revolves around the notion of workspaces, so native support for that changed what was possible,” says Priya. “Being able to establish one-to-many and many-to-many relationships without a consultant is what keeps the data clean and legible.”
Kelvinlabs uses Perigee's native Conduit integration to keep the CRM current with the latest product data.
“We use Conduit to pull in all of our workspace information and associate it with the right companies and people. We've also been importing intent signals — usage patterns, failed jobs, quota pressure — so we can reach out before an account goes quiet. Being able to map and transform that data has moved our retention rate meaningfully,” says Priya.
Automate, iterate, and accelerate
Using Automations, Kelvinlabs has built a set of workflows on top of Perigee to power lead generation.
“We're constantly using Perigee to find accounts that are ready to upgrade, or customers who've outgrown the plan they're on, and to give those people a better first conversation,” says Priya. “Using workflows, we've been able to turn that process into an engine rather than a weekly meeting.”
As a result, Kelvinlabs has built a range of automated workflows in Perigee that are deeply integrated with the rest of the stack. By drawing on Perigee as their single source of truth, the team can proactively engage with warm leads in a far more efficient way.
“Perigee's data enrichment capabilities make it straightforward to identify potential leads. We have a number of workflows to filter leads by enriched attributes, such as funding_stage and seat_count, which lets us conduct highly targeted and customised outreach,” says Priya.
By automating these processes in Perigee, Kelvinlabs can prioritise more strategic initiatives to accelerate their growth.
The center of their GTM stack
In addition to the native Conduit integration, Kelvinlabs has worked with the Perigee API to build some of their own integrations.
“Perigee's API documentation is incredibly thorough and intuitive, which made it easy to write integrations for other tools in our stack,” says Priya. “For example, we set up an integration with our email campaign platform, Sendline, which lets us trigger marketing campaigns based off people we've identified in Perigee.”
Kelvinlabs has also linked Formwork, a form builder, to their Perigee workspace. “Using the webhook from Formwork, I can instantly send events to Perigee, which ingests the people who have completed the form into a list. It means it takes almost no work to act on this data,” says Priya.
Scaling up alongside Perigee
Kelvinlabs is playing a key role in driving their future growth. “The product is continually evolving — we can see distinct improvements month by month, which you just don't get with the incumbents,” says Priya.
For Kelvinlabs, Perigee has been influential at getting the team on the same page and accelerating their PLG motion.
“To see the whole team, including the co-founders, be excited about a CRM is not something I expected,” says Priya. “Everyone is noticeably bought into it, and actually happy to look at data in Perigee, add notes, build lists, and interact with the platform as a whole. I've never seen people so engaged in a CRM.”
“We're looking forward to building more workflows on top of Perigee and integrating it more deeply with the rest of our stack. I'm excited to see the product evolve in the years to come, and have faith in Perigee's ability to scale with Kelvinlabs as we grow.”
