Our Investment in Seldon — Democratizing AI in Enterprises
As economies across the world are recovering from the impact of Covid-19, one thing is more certain than anything else — businesses, small to large, are embracing AI and machine learning in their core business processes. Even before Covid-19, many enterprises, from global pharmaceuticals looking for an edge in drug discovery to hedge funds on wall street optimizing their portfolio models, have been increasingly relying on machine learning models as an integral part of their decision-making process.
“Large enterprises have begun a shift to standardizing their compute infrastructure to common technology platforms in order to leverage economies of scale in tooling, knowledge and policy application across silos, and efficiencies in software resource management”
Seldon’s platform simplifies the process of managing, deploying and monitoring ML models at scale in enterprises with complex heterogenous environments. It removes the friction points that IT Ops teams typically have to deal with when business users want to incorporate machine learning but the existing IT infrastructure isn’t set up to support this. To overcome this problem, enterprises typically had to build teams of experienced ML engineers and pair their up with IT operations. Seldon removes this friction and eliminates the pain points by standardizing the process of ML model management incl. testing, deployment and monitoring. Its cloud-native platform is currently used by fortune 100 companies across Europe and US incl. the likes of Capital One and Nasdaq. On average, customers who use Seldon were able to reduce the time required to get their ML models in production from three months to a few hours.
“ML and AI workloads have traditionally been run on special-purpose infrastructure on account of their particular requirements — high CPU, large amounts of memory and potentially hardware GPU access. In some organizations this has led to a proliferation of home-grown clustering and resource management in silos separate from more traditional application workloads”
When I spoke to Alex and Clive earlier this year, it was clear to me that they are solving for a complex yet fundamental problem faced by enterprises that are keen to leverage the power or AI and ML. After spending two decades at Cisco, EMC and SunGard, I know way too well the complexities that IT teams face on a regular basis to keep up and accommodate for business needs. Using Seldon, enterprises can not only eliminate the pain that IT teams face on a day-to-day basis but also helps them bring closer to business users.
We are excited to partner with Alex and his team on their journey to build the next-gen ML operations platform that will accelerate the democratization of AI across enterprises.
Vin Lingathoti is a partner at Cambridge Innovation Capital and focuses on investments in AI/ML, Cybersecurity, Cloud, IoT and DevOps. Before joining CIC, he led venture investments and acquisitions for Cisco in London and Silicon Valley. Earlier in his career, he held product management and operations roles at EMC and SunGard. Vin is a software engineer and has spent over a decade in Silicon Valley working with early-stage tech companies.
