Almost every company operates on a digital platform to service their customer needs. Whether it’s an ecommerce business or an enterprise running internal applications, they use the same technologies underneath. The smooth functioning of these technologies is critical; any disruption can directly affect the business’s performance and continuity.
Enabling the frictionless operation of these technologies, unSkript, a tech infrastructure health intelligence platform, is announcing a $3.75m pre-seed funding round to identify and mitigate issues in the software that powers the world.
Fragile software systems that power the world are running atop large computing clusters, often using new technologies such as Kubernetes on both public clouds and on-premises environments. These complex environments are held together by specialist software engineers in DevOps and Platform Engineering teams who often work late into the night solving complex problems to keep society running. These teams play a vital role in identifying and mitigating issues in the software that powers the world.
Traditionally, these teams have depended on undocumented knowledge and textual knowledge bases (aka runbooks or playbooks) to support them when navigating complex incidents. However, if these runbooks are misinterpreted or outdated, the engineers will often be left on their own to debug and resolve these challenges during live incidents.
unSkript leverages Generative AI to assist these teams to proactively and automatically detect and resolve potential infrastructure health issues and production downtime; mitigating the risk of human error and decreasing developer burnout.
unSkript solves this using a Large Language Model (LLM) agent that learns from existing knowledge bases and user activity to understand procedures and match them with problem signatures. With unSkript, companies can shift from manual learning and execution of procedures to an AI-driven approach where the Generative AI agent assists in triaging and suggesting the next-steps, thereby cutting the risk of human error and burnout. All health-check scans or test are available on opensource here: https://github.com/unctl-sh/unctl
“unSkript ships hundreds of pre-built failure signatures (aka health checks) that a team can directly leverage. These signatures span many cloud-native technologies such as kubernetes, databases, messaging, and IAAS cloud. Along with the signatures, we ship an LLM agent to learn new signatures and provide the next steps on a current problem. As a result, we can reduce troubleshooting time by up to 12x and resolution time by up to 25x” said Abhishek Saxena, CEO of unSkript.
Abhishek Saxena
“Monitoring and troubleshooting our customer deployments was a significant challenge for us on our growth journey. unSkript’s approach of using AI as a Co-pilot has been instrumental for us to scale and be successful” said Himanshu Shukla, CEO of LightBeam. unSkript brings on-call teams into the post-LLM world where health checks, troubleshooting and fixes are all done by AI. As the product and the organization scale, the number of deployments and repetitive problems increases exponentially, making individual management impractical.
The company has raised $3.75m in pre-seed funding so far from leading early-stage technology investors such as Westwave Capital, First Rays Venture Partners, Scribble Ventures, Zero Prime Ventures, and notable angel investors.
“In today’s rapidly evolving tech landscape, the complexity of IT environments is escalating with the integration of cloud services and distributed systems. Over time, as more deployments get added due to scale, compliance, and customer requests, the repeatability of the problems tends to increase. This is where a scalable and intelligent solution like unSkript emerges as a pivotal solution, accelerating business agility while significantly reducing the risk of human error and downtime,” said Gaurav Manglik, Partner, Westwave Capital. The unSkript team, given their vast experience, is uniquely positioned to solve this growing problem.”
