Upwind Security

When GenAI tells you what happened in an incident, what it’s going to do, and it’s right again and again, by the 81st time, you let it run. Because it earned your trust.   That’s the future we’re building at Upwind. 💪   And over the past month, you’ve seen that come to life.   We created Choppy AI for security teams to quickly understand their environment and take action using natural language.   And with Unified AI Protection, we've brought AI posture, inventory, behavior tracing, and validation into the same runtime-first workflows you already use.   Not just answers. Real context. Real trust. That’s how we get to agentic AI.   Up & Upwind! 🏄♂️ Amiram Shachar

Most tools start with CVE severity. Upwind starts with what’s actually happening inside your environment. Last week, we launched Choppy AI, a natural language assistant that turns plain English into precise, runtime-aware queries. In the demo, we asked: “Which Kubernetes workloads are publicly exposed and have critical runtime-based vulnerabilities?” Here’s what happened next:  1. Choppy built a query using real runtime context with no DSL required  2. It pulled in internet exposure data based on live ingress traffic  3. It filtered for CVEs that are actively loaded and exploitable at runtime  4. It prioritized results based on what’s truly at risk, right now  5. We saved the query and converted it into a policy tied to our internal audit framework This is how Upwind helps you go from question to action in seconds, powered by an inside-out view of your cloud. Watch the full demo below. Up & Upwind! 🏄♂️ Ben Engel

We just rescued a new puppy late last year, and it’s brought so much joy into our already busy home. So maybe I’m being a little sentimental when I look at Upwind Security’s new "Choppy" as my loyal cloud security best friend. Upwind’s UX is already one of the best in the cybersecurity industry, but what really excites me is where it’s heading. The future of UX is personalization, and Choppy feels like a real step in that direction. It helps me (and our team) get to the context that matters most, faster. Excited to see Choppy grow with Upwind. It’s the new puppy that actually helps clean up the messes instead of creating them 🐶☁️

Our CEO, Amiram Shachar recently shared his winning AI security model on Amazon Web Services (AWS) Security LIVE! 🎧   "AI security should not be a tool on its own. It can consume a lot of the context and shared knowledge your cloud security platform already has, like network topology, API traffic, SBOM, and vulnerabilities – and make it specific to AI."   At Upwind, we make this possible by securing your runtime cloud environment, giving AI the clarity it needs to act fast and smart.   Up & Upwind! 🏄♂️

We’re giving away 10 Upwind YETI dog bowls 🐶 In honor of our latest launch, Choppy AI, named after our office dog, Choppy. To enter: 1. Post a photo of you and your dog 2. Share a few thoughts on why dashboards ≠ better security 3. Tag Upwind Security Entries close Friday, January 16. Winners will be selected by our CSO, Rinki Sethi, and notified via LinkedIn DM. One entry per person. Let’s see your best dogs, and your best takes. Up & Upwind! 🏄♀️ This promotion is not sponsored, endorsed, or administered by LinkedIn.

GenAI is moving fast and so are the risks. Do you know how it’s being used across your cloud? Upwind Security gives security teams the visibility and control they need to stay safe and compliant: • Map every workload talking to GenAI platforms like OpenAI, Bedrock, and Vertex • Flag misconfigurations like exposed models or weak access controls • Detect suspicious usage and policy violations in real time • Get alerts when sensitive data (like PCI or internal docs) is sent to GenAI services With Upwind Security’s runtime-first CNAPP, you can embrace GenAI without exposing your environment. Stay ahead of compliance risks, prevent data leaks, and keep usage aligned with policy. Up & Upwind! 🏄♂️ Bar Klinghofer

Last week, Upwind Security announced a new feature called Choppy AI, which immediately caught my attention. Choppy is not just another LLM added on top of the UI. It is tightly integrated into the Upwind ecosystem, and that integration really shows. Choppy appears in multiple areas across the platform and delivers a consistent experience that genuinely reduces friction and improves productivity. One area we have been investing more in is custom configuration rules, which better reflect our organization’s specific needs and align with increasingly complex security certifications and requirements. What stood out to me about Choppy is its strong support for this custom configuration workflow. Instead of writing Rego rules from scratch, we can describe the intent we are aiming for, and Choppy generates rules that already understand Upwind’s data model and capabilities. For teams dealing with complex security requirements, this shortens the path from intent to implementation in a very practical way. #CloudSecurity #securityai #choppy 🐶

Last week, Upwind Security announced a new feature called Choppy AI, which immediately caught my attention. Choppy is not just another LLM added on top of the UI. It is tightly integrated into the Upwind ecosystem, and that integration really shows. Choppy appears in multiple areas across the platform and delivers a consistent experience that genuinely reduces friction and improves productivity. One area we have been investing more in is custom configuration rules, which better reflect our organization’s specific needs and align with increasingly complex security certifications and requirements. What stood out to me about Choppy is its strong support for this custom configuration workflow. Instead of writing Rego rules from scratch, we can describe the intent we are aiming for, and Choppy generates rules that already understand Upwind’s data model and capabilities. For teams dealing with complex security requirements, this shortens the path from intent to implementation in a very practical way. #CloudSecurity #securityai #choppy 🐶

AI in cloud security shouldn’t be a black box. We built Choppy AI to give teams what they’ve been missing for too long: not just answers, but explainable ones. Grateful to Enterprise Security Tech for diving into how Choppy AI flips the script and why transparency matters more than ever. It’s one part of a bigger mission: bringing visibility, trust, and actionability to every layer of AI in the cloud. ☁️ If you’ve worked with AI in your security workflows, what’s helped you trust the output and where do you still want more clarity or control? Drop your thoughts below. 👇

Choppy AI is out in the wild and already sniffing out real risks. 🐶 Yesterday, we unleashed a new way to think about cloud security: Not just detecting risk, but actually understanding it. Because security teams aren’t struggling with signal. They’re drowning in alerts, dashboards, and noise. So what does that look like in action?  1. Ask questions in plain English  2. Turn intent into real policy and logic  3. Investigate and prioritize with context It’s clarity at cloud scale. Built into the way your team already works. If you're wondering where the name came from... Choppy AI is named after our Shih Tzu mascot who’s great at fetching what matters – so is our AI. Ask a question, get the right answer with context you can trust. See how Choppy AI brings clarity to your cloud security 👇 Up & Upwind! 🏄♂️