Welcome to Snowflake’s Startup Spotlight, where we learn about awesome companies building businesses on Snowflake. In this edition, find out how Evan Powell, founder and CEO of DeepTempo, is harnessing AI alongside a team of skilled security experts to protect the digital world from increasingly sophisticated cyberattacks.
DeepTempo defends against increasingly sophisticated and relentless attackers through deep learning that is proven to find advanced attacks that earlier generation systems miss while reducing costs thanks to low false positives and our partnership with Snowflake.
I want to protect the digital world, harness collective defense and deep learning for cybersecurity and take back the initiative from attackers.
This is my sixth enterprise-focused startup. The more I do this work, the more I think that success comes down to people. Building relationships of trust with a team and an extended network of believers — and even critics — is all-important. From trust come the best hires, and also the best customers and partners.
We are AI native. Everything we do, including our core product, is based on AI and specifically deep learning and Log Language Models (LogLMs). AI enables us to reimagine customer workflows and solve challenges that have long plagued everyone in cybersecurity. With AI, we can also move much faster than earlier-gen organizations.
We are all under attack. Increasingly, innovative professional attackers are working their way around and through our thicket of cybersecurity defenses, often with the aid of AI. Aging rules-based systems are failing to detect attacks effectively. Meanwhile, the costs of cybersecurity licensing, operations and maintenance are always rising, even as these systems prove ineffective and verge on obsolescence.
Our application, Tempo, is available as a first-of-its-kind Snowflake Native App on Snowflake Marketplace. Tempo helps customers identify security incidents and analyze their scope and severity. The app was pretrained using enormous quantities of security logs and is particularly focused on the pattern of events, including relative and absolute time. It has been shown to be extremely accurate, with a low false positive and false negative rate. The Bank of New York is an important supporter and we have completed their Ascent program and an extensive POC.
Most importantly, at DeepTempo, we have assembled a unique mix of deep-learning experts and data scientists, many of whom have worked as security operations center (SOC) analysts and detection engineers. Our technical depth is matched only by the depth of our experience running security and other technical operations at scale.
We chose Snowflake because it lets us focus on helping customers protect their enterprise instead of getting bogged down in gathering, approving and transferring data into our system for analysis.
The Snowflake Native App Framework has helped us massively accelerate our flywheel. Instead of trying to move all of the data into a black-box security system to analyze it, our approach brings the “brain,” or intelligence, to the data. The Snowflake Native App paradigm allows us to piggyback on the governance work of Snowflake, which means security teams can start testing and using the tool right away, as opposed to spending four to five months on internal approvals. The impact is enormous: They can seamlessly integrate it into their existing tools, such as Anvilogic or Splunk, in just a few minutes.
It also reduces vendor lock-in, giving enterprises the freedom to use Snowflake as their primary platform while selecting the best tools for their needs. This combination of flexibility and speed make a massive impact. Many CIOs do actually think about lock-in — after all, they’ve supported their organizations in adopting open source. I think the benefit of Snowflake and Snowflake Native Apps’ ability to decrease vendor power while increasing best-of-class adoption is underappreciated.
We created something called Log Language Models, or LogLMs. They are better at spotting real security problems without raising too many false alarms. They start working right away, adjusting to new setups in hours or days instead of months with prior generation ML. And they’re easier to understand, so the security team is always aware and in control of what’s happening. Using Snowflake to store and manage data for these models helps save a lot of money and time.
There’s nothing else like the Snowflake Native App environment for enterprises today. LLMs and AI can outperform traditional, proprietary SaaS systems, giving AI-native companies an enormous productivity boost. The Snowflake Native App approach lets you focus on being the best at your specific use case, rather than wasting resources reinventing the wheel and building your own data infrastructure.
I would tell others to look at new use cases and business models made possible by Snowflake Native Apps. Improving existing use cases is just the beginning — what can your users do that was previously impossible, and how could it change their business? Think about second-order effects and long-term impact.
Discover how DeepTempo is transforming cybersecurity with cutting-edge deep learning at deeptempo.ai and explore its dynamic collaboration with Snowflake at deeptempo.ai/snowflake. You can also try out the Tempo app for yourself.
As businesses and individuals increasingly rely on cloud services for storage, collaboration, and computing power, the importance of securing cloud […]
Early enterprise adopters of generative AI have made it clear that a robust data strategy is the cornerstone of any […]
Snowflake’s Accelerate 2025 virtual event series offers a crucial opportunity for public sector and healthcare and life sciences organizations to […]
Our services help both the big and small organisations in the identification of security breaches and provide outclass solutions to prevent any potential harm.