Tines’ Post

In today’s Friday Flows, Conor Dunne, from the Tines Labs team, walks us through a new story using AI to create cases and act on CrowdStrike alerts. As is the case with many alerts, there’s a lot of information, but it’s not always very clear. He first uses AI to simplify & normalize the data. Once that is done & a case is created, we can also use AI to act as a security analyst and respond with one of four actions: 🙅♀️ Suspend a user account 🔐 Isolate a host ❌ Block a URL 🔔 Alert the security team using PagerDuty In this demo, the AI was confident enough to block the URL, so it took action. If the confidence is not high enough, it will provide a recommendation but allow the analyst to take action manually. I love this example because it’s a natural evolution of stories that Tines users have built for years: take an alert, enrich it, create a case, and help an analyst act. But with the developments in AI, it’s now much easier to parse the information & act with fewer manual steps. To see this story and more AI-incorporated workflows, check out the new section of the Tines Story Library here: https://lnkd.in/exk__G7U

🚀 Coming out of #MRMC24 this week, the hottest topics were around battling marketplace risk using AI and how to reimagine fraud & risk prevention. 🚀 It's timely for me to share this Ask the Expert segment featuring the incredible T. Scott Clendaniel, a top voice in AI on LinkedIn! Today’s interview is a love letter to all my fellow data nerds. 📊💡 Listen to Scott as he will demystify GenAI's Wins, Losses, and Opportunities, sharing insights from his 30+ years in the DS/AI industry. With experience consulting for giants like Wells Fargo, Cisco, and Mercedes, and soon joining Gartner as their Director in Data, Analytics, and AI, Scott's knowledge is unmatched. Tune in to discover the latest trends, from neural networks to Deep Learning, and learn about the real impact of GenAI on democratization, business efficiencies, and Trust & Safety. Don't miss out – Scott's expertise will blow your mind! 🤯 Looking to enhance your ML/AI initiatives? Connect with me to learn how Vaco can help with human review of AI workflows to battle fraud and bolster Trust & Safety. Vaco helps companies define their QA rubric, achieve >99.5% quality in less than 6 weeks, while designing AI policies and implementing a knowledge base that increases their training & ML model accuracy. This is where AI proves its innovation by protecting brands and users. As we know well, automation and machine learning are the tried and true solutions we can depend on. They also still require human oversight & audits. #AI #DataScience #GenAI #Innovation #TrustandSafety #ExpertInsight #Vaco #AITrends #MachineLearning #HumanReview #FraudDetection #TrustTuesday #RiskandCompliance #PolicyChangeManagement

Adding a new security DLP policy can be very difficult for reasons including, staffing constraints, expertise, complexity, and operational requirements. Innovative security providers are using AI to bridge this gap. Here is an example no-code policy request into Skyhigh, where AI would interpret and generate expressions to be used for a DLP policy. "Build me a regex based rule to protect all my credit card formats including VISA, MasterCard and even including AMEX card formats" See below for more detail.

Great experience going to nPlan's AI day on Wednesday. Have a look at Alan's summary below. I recommend going to the next one in May. They demonstrated their 💼 portfolio product, which I hadn't seen before. It's worth watching the replay for this alone. A good demo of the 💡 Insights platform, which has been subject of previous posts. Some of the user tips I picked up: 1. Barry is doing a lot of the work under the hood, it labels it as ' Powered by Barry'. 2. you can focus on the critical path for any milestone, not just the final one 3. with the risk functionality, you can ask Barry for impact scores. 4. Barry looks for opportunities as well as risks if your register includes them 5. ask Barry to summarise similar activities to an activity of interest by 'taking from the wbs', which is your corpus of historical schedules. 🕸 Keenly looking forward to the roll-out of GraphGen over the next few months: it will go from text to schedule graph and vice versa, as well as from schedule to improved schedule. There's a lot more than that in the replay. 🖥 Keenly looking forward to next steps of the ongoing roll-out of their ICE-LM language model into the services that Barry provides (Insight and Portfolio). It will also power a lot of the functionality in Graph Gen, as well as be able to generate possible project scenarios/eventualities for you to stress test your planning. Great to see a team in the project space pushing for definitive advances at least every six months. We need more of that.

Deploying a Machine Learning Model with NOTO! We're thrilled to bring you the latest update from our machine learning journey with NOTO. In this video, we're diving deep into the deployment of our trained model within the ruleset. Here's a quick overview of what you can expect: 🔍 Reviewing Model Characteristics: Before proceeding with deployment, we take a moment to evaluate the key characteristics of our trained model. From assessing feature importance to analyzing performance metrics, we ensure our model is primed for success. 📈 Performance Analysis: Delve into the intricacies of the Receiver Operating Characteristic curve, gaining insights into our model's ability to identify fraudulent transactions while minimizing false positives. We fine-tune parameters to optimize real-world performance. 💡 Deployment Process: Witness the seamless integration of our trained model into the ruleset. From updating model versions to defining trigger conditions and actions, we demonstrate the step-by-step process of deploying our machine learning solution for effective fraud detection and investigation. 🔒 Ensuring Accuracy and Efficiency: With meticulous attention to detail, we prioritize accuracy, efficiency, and reliability in our deployment strategy. Leveraging the power of NOTO, we empower organizations to stay ahead of emerging threats and safeguard their operations. Stay tuned for more insightful updates as we continue to explore the limitless potential of AI together! #MachineLearning #NOTOPlatform #FraudDetection #AI #DataScience #Innovation

Insightful context as DC adapts to upcoming technological advancements - learn more at our upcoming virtual roundtable!

On Halloween Eve 🎃, the Biden administration released a major executive order (EO) on AI policy. Nothing too spooky 👻 ...but a lot is happening. Here are a few takeaways: 👉 NIST at the center: The EO empowers NIST to develop standards around AI. This makes sense; NIST has a history of collaborating with the tech industry and updates a widely-used a Cybersecurity framework. 👉Speed: The speed at which AI policy is developing is phenomenal. A year ago, this was barely on the regulatory or legislative agenda. 👉Stakeholders at the table: For those paying attention, the EO wasn't a complete surprise. The EO aligns with what I've heard in conversations for months. Since mid-year, the administration has been in talks with industry and advocacy groups to blaze a path forward. And the EO builds upon previously announced efforts, including a blueprint for an AI bill of rights and voluntary commitments from 15 tech firms. 👉AI will affect everything: The sheer breadth of the EO points to how AI will affect every industry -- along with every agency that writes rules. 👉Industry wants thoughtful sensible AI regulation: The industry engagement so far points to what I've heard on the ground: those in tech want to be proactive in shaping sensible regulation. The last thing they want are bad actors to cause a massive AI backlash that results in expansive regulation down the road. And what they really don't want are AI policies with unintended consequences. 👉A geopolitical (and hyperlocal) policymaking race: The U.S. isn't alone in policymaking. Tomorrow, UK PM Sunak hosts an AI Safety Summit. The G7 countries announced a International Code of Conduct for Advanced AI Systems earlier in the week. And there are dozens of states that have proposed their own policies. Virtual Roundtable Discussion 11/7 🖥 This is just a quick take. If you track AI policy for work, next week National Journal is hosting an exclusive AI policy virtual roundtable discussion where we will cover more. To register / request an invitation to the roundtable visit https://lnkd.in/ezg6Rhq6

Did you know that Splunk is an excellent base for Ai/ML? And the Ai/ML piece is for FREE 😳 We will manage the normalization and handle the volumes of data needed, and we do that really really fast. In this e-book you will get: - An introduction to AI and ML in observability - The Splunk AI/ML portfolio (lots of goodies) - How to start your AI/ML project (easier than you think) - Cross-industry use cases for Splunk observability Check out our "cook-book" on how to get started: https://lnkd.in/dx2Wyq8z

Wendi Whitmore SVP at Palo Alto Networks Unit 42 sat down with Liz Claman to discuss the power of #AI in accelerating and automating government agencies defenses mechanisms against emerging cyberthreats. Some takeaways from this quick 4 minute video -- AI is a double edge sword. Can be used for good and can be used for harm. What we are seeing in 2024 - speed, skill and sophistication. Our prediction is the volume of activity. Threat actors can leverage written communication to get a human to conduct in a behavior to click a link or answer a phone call. To learn more - watch this video via Fox Business Network. https://lnkd.in/eTvBfbB6

I am excited to announce our upcoming webinar on one of the hot topics! Join me and my colleagues today to learn how to successfully pilot Generative AI in your organizations. We are going to cover several topics like security & risk, use-cases, skills, responsible AI, build vs buy options and many more! Here's where you can register: https://lnkd.in/ebA397Cm Leinar Ramos, Avivah Litan and Radu Miclaus #GenerativeAI #Gartner