About
Articles by Bipul
Activity
76K followers
-
Bipul Sinha posted thisWe coined the term “agentic cyber resilience” to replace the term “cybersecurity.” Cybersecurity applied to the assumption that human attackers would go after human-run businesses. This assumption is now wrong. AI is attacking at machine speed. Agents are running business processes at machine speed. The old architecture won’t withstand this new world. To have confidence in your ability to run digital business and deploy agents, you need to control both external AI attacks and internal agentic risks. Agentic cyber resilience addresses both. The world of cybersecurity is gone. The world of agentic cyber resilience is here. It’s time to rethink your strategic posture.
-
Bipul Sinha posted thisAI is our generation’s most powerful technology. When it's developed responsibly, it comes with safeguards, red teams, and oversight. Governments are involved. The risks are debated openly. But powerful technology doesn't stay contained. Frontier AI capabilities are now open-weight, freely available, and in the hands of anyone who wants them. No restrictions. No accountability. The threat to enterprise data is no longer theoretical. It is autonomous, accelerating, and it doesn't ask permission. That's why cyber resilience isn't optional. It never was.
-
Bipul Sinha posted thisThe number one job of a founder and CEO: stack the next S curve. Once a first product sees success, it’s time to start building the second and third product foundation. With AI compressing the scale and decay of product cycles, you have to be particularly persistent about this process. A strong portfolio of non-consensus ideas. Products continuously tested against the market. Scaling fast. And when the run is done, doing it all over again. One S on top of another. Ad infinitum. That's how you build a business that lasts.
-
Bipul Sinha reposted thisBipul Sinha reposted thisThe biggest news today in AI – while doing research into frontier models for cybersecurity, OpenAI's test model broke containment and compromised the infrastructure at Hugging Face. CISOs often ask me what Mythos and other frontier-class AI models mean for the future of cybersecurity. I think this is probably the clearest indication. The lesson for me isn't "AI is scary." It's that model capability now outpaces the controls around it — and the same is about to be true of the controls around most of our organizations. The flip side: these capabilities cut both ways. The labs just proved AI can find and chain vulnerabilities faster than humans could flag — which means it can do that for defenders first. That's the whole game now: use AI to close the gap between breach and recovery, before an attacker (or an agent) opens it. And finally – you can't out-run machine-speed exploitation with prevention alone. You have to assume breach and have a way to recover just as quickly as you're compromised. Details in their official statement: https://lnkd.in/gNstTdUt
-
Bipul Sinha shared thisYou cannot wish cyber risk away. It’s part of your life. Breaches will happen. So the question is: how are you going to manage when they do? With resilience. Resilience is a combination of two things. It’s the understanding of the risk (so you’re prepared) and the ability to remediate and restart services when breaches happen. You need both. A company that can detect and prevent, but cannot recover quickly, is not resilient. Complete resilience means withstanding an attack and keeping your business going. That’s the standard every company should be building toward. Risk + remediation = resilience.
-
Bipul Sinha posted thisAI agents are the nuclear power of the digital world. Massive opportunity, massive risk. They can offer companies 100x their productivity. But if they are compromised, they can do 10x more damage…in 1/10th of the time. That's why most AI pilots never reach production. The opportunity is clear. The guardrails aren't. So, how can you gain the productivity without incurring the damage? As with a nuclear power plant, you need guardrails. And agentic AI needs dynamic guardrails. Not static rules - dynamic guardrails that understand intent and stop bad actions before they start. AI agents have great capacity for good…and great capacity for evil. To harness their power for good, move beyond monitoring. Understand intent. Build dynamic guardrails. To capture the upside, you have to contain the downside first.
-
Bipul Sinha shared thisAnthropic’s Mythos is a milestone model. It can chain vulnerabilities and find toxic combinations that result in intrusion and breach simultaneously. In other words, it’s changing the entire game of cybersecurity. With no time to detect anything, you have to ensure you have machine speed resilience - what I call agentic resilience. Attacks will be continuous, so your recovery needs to be continuous. That’s the world Mythos has ushered us into. If you’re still building security postures for a pre-Mythos world, you’re behind. The era of agentic resilience is here.
-
Bipul Sinha shared this¡Vamos Spain! Last week, Rubrik partnered with Microsoft to host our CXO Resilience Summit in Madrid. The room was full of leaders asking the same question: in an AI world, what does resilience actually mean? Our answer: recovery, integrity, and control. Not as three separate disciplines. As one requirement. AI is rewriting how businesses operate and how attackers operate. The organizations that win will be the ones who can recover fast, trust their data, and stay in control of both. Grateful to Microsoft for the partnership, and to every CXO who spent the day building the future with us.
-
Bipul Sinha posted thisAt Rubrik, we do everything in service of Rubrik. So we’re always seeking the truth. And conflict and disagreement are the byproducts of that truth-seeking. If everyone is being polite and agreeable, we’re not getting the truth. And without the truth, we’ll be making decisions based on incomplete information - no matter how many smart people are in the room. That’s why we debate at Rubrik. First, because we know that the best way to avoid blind spots is to invite all the brains into the game. And second, because our agenda is the company's agenda. Not personal success or ego. So we encourage debates and discussion, never hypocritical politeness. Only if everyone shares their opinion - an opinion with Rubrik’s success at the forefront - can we get to the truth. And if those opinions come into conflict with one another, all the better. That means we’re leaving no stone unturned. Thus, the conclusion we come to will be the strongest one for the betterment of Rubrik.
-
Bipul Sinha liked thisBipul Sinha liked thisThirty years ago, NYSE and CNBC made a shared commitment: to bring the heartbeat of the markets to the world. That commitment has never wavered. Today, we celebrate three decades of partnership built on trust and transparency. Congratulations to President KC Sullivan and the extraordinary team at CNBC - those who deliver clarity when it matters most. This partnership is also a powerful reminder of freedom of speech and the free flow of information. Transparent markets don't exist in a vacuum. They depend on journalists who can ask hard questions. CNBC has been that steadfast, reliable voice for three decades, and the NYSE is proud to stand beside them. Here's to many more years of ringing bells, breaking news, and telling the profound story of our global markets.
-
Bipul Sinha liked thisBipul Sinha liked thisToday, I'm excited to share that I've joined 𝐓𝐡𝐞 𝐖𝐚𝐥𝐭 𝐃𝐢𝐬𝐧𝐞𝐲 𝐂𝐨𝐦𝐩𝐚𝐧𝐲 𝐚𝐬 𝐕𝐢𝐜𝐞 𝐏𝐫𝐞𝐬𝐢𝐝𝐞𝐧𝐭, 𝐈𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧 𝐒𝐞𝐜𝐮𝐫𝐢𝐭𝐲 𝐒𝐭𝐫𝐚𝐭𝐞𝐠𝐲, 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠, & 𝐀𝐫𝐜𝐡𝐢𝐭𝐞𝐜𝐭𝐮𝐫𝐞․ Disney has long represented the very best in creativity, storytelling, and innovation. It's a privilege to join a company whose stories, experiences, and technology have inspired generations and continue to create moments of wonder for people around the world. Throughout my career, I've believed that security is most impactful when it enables the business, helping teams innovate with confidence, responsibly adopt emerging technologies, and strengthen the trust of customers, employees, and partners. I'm excited to collaborate with the talented teams across Disney as we continue building a strong security foundation that supports creativity, innovation, and the responsible use of AI across the enterprise. My sincere thanks Andrew Kirkland and the leadership team for this opportunity. I'm grateful for the warm welcome and look forward to learning from my new colleagues, building strong partnerships, and contributing to the important work ahead. 𝗔𝘀 𝗪𝗮𝗹𝘁 𝗗𝗶𝘀𝗻𝗲𝘆 𝗳𝗮𝗺𝗼𝘂𝘀𝗹𝘆 𝘀𝗮𝗶𝗱: ❞𝗪𝗲 𝗸𝗲𝗲𝗽 𝗺𝗼𝘃𝗶𝗻𝗴 𝗳𝗼𝗿𝘄𝗮𝗿𝗱, 𝗼𝗽𝗲𝗻𝗶𝗻𝗴 𝗻𝗲𝘄 𝗱𝗼𝗼𝗿𝘀, 𝗮𝗻𝗱 𝗱𝗼𝗶𝗻𝗴 𝗻𝗲𝘄 𝘁𝗵𝗶𝗻𝗴𝘀, 𝗯𝗲𝗰𝗮𝘂𝘀𝗲 𝘄𝗲❜𝗿𝗲 𝗰𝘂𝗿𝗶𝗼𝘂𝘀... 𝗮𝗻𝗱 𝗰𝘂𝗿𝗶𝗼𝘀𝗶𝘁𝘆 𝗸𝗲𝗲𝗽𝘀 𝗹𝗲𝗮𝗱𝗶𝗻𝗴 𝘂𝘀 𝗱𝗼𝘄𝗻 𝗻𝗲𝘄 𝗽𝗮𝘁𝗵𝘀.❞ Next week, I'll be attending Black Hat USA and DEF CON in Las Vegas, where I'll have the opportunity to reconnect with friends and colleagues from across the cybersecurity community. I'll also be there with my son, Matthew Harris, who is building his own career in cybersecurity. It's become one of our favorite traditions, and I'm looking forward to another week of learning, exchanging ideas, and catching up with so many people who make this community so special. If you'll be in Las Vegas at summercamp, I'd love to connect. #securethemagic #AIsecurity #SecurityLeadership #Hackersummercamp #Cybersecurity #CastMember
-
Bipul Sinha liked thisBipul Sinha liked thisThe model market is changing faster than most enterprise AI architectures can absorb. New frontier models arrive every few months. Open-weight models are improving on cost, portability, and deployment flexibility. Frontier models will continue to push the edge of what AI can do. Open-weight models will expand choice and make more workloads economically viable. Enterprises need both. This is why Glean signed the Open Weights and American AI Leadership letter. It’s a position we’ve held for a while now. The competitive advantage in enterprise AI won’t come from choosing one winning model. It will come from giving employees the right intelligence for each task (using frontier models where capability matters and open-weight models where cost, control, portability, or deployment flexibility matter) without forcing your organization to rebuild its AI stack every time the market changes. But that only works if your AI architecture can evaluate, route, govern, and replace models without rebuilding the system around them. That’s what we’ve built at Glean: a shared system for enterprise context, permissions, policy, observability, and model routing, so customers can take advantage of a changing model market without re-architecting every time it moves.
-
Bipul Sinha liked thisBipul Sinha liked thisI am excited to announce my new role as SVP, Partners & Technology at Mistral AI. After four fantastic years at Confluent, I am excited to be building global partner ecosystems at one of the leading frontier labs in the world. At Mistral, I will lead the Partner Business, focusing on Compute, GSIs, ISVs, MSPs, and NCPs. I am grateful to everyone at Confluent who made the last four years so special. I am energized by what comes next. Let's build. 🚀 #MistralAI #EnterpriseAI #Partnerships
Experience & Education
-
Rubrik, Inc.
View Bipul’s full experience
See their title, tenure and more.
or
Other similar profiles
Explore more posts
-
Paul Drews
Today marks a milestone for Salesforce Ventures, and a defining moment for enterprise security. Alphabet’s acquisition of Wiz has officially closed. Wiz has been a cornerstone of our portfolio since May 2021, when Nowi Kallen, Kartik Gupta, and team first invested in their Series B at a $1.5B valuation. We saw the potential early and invested deeply, backing the team through their Series C, D, and E rounds. From a $1.5B Series B to a $32B acquisition, one of the largest in enterprise tech history, the Wiz team’s execution has been nothing short of extraordinary. This outcome reflects our mission at Salesforce Ventures: identifying the companies that define the next era of technology and staying behind them for the long haul. But the real credit belongs to the entire Wiz team on the ground. Beyond the technology, it was the relentless grit and day one mentality of Assaf Rappaport and the team that turned a vision into a $32B reality. Huge congratulations to the entire Wiz team on this historic achievement.
-
Mike Duboe
From Measurement to Orchestration, with Pranav Piyush from Paramark.com In this week's Intelligent Marketer ep, Rishabh and I dug into the evolution of marketing measurement in a world where direct, click-based attribution has broken. The smartest marketing teams are building systems that triangulate incrementality testing + MMM results. Pranav discusses how getting this "central brain" right is a critical foundation for AI-driven orchestration & execution of an end-to-end marketing roadmap. Topics covered: - Policies, not dashboards: how to decode decisions as machine-readable rules that turn insights into actions - Confidence-weighted budgets: how to use error bars in lift/MMM results to drive budget reallocation - Latency-aware orchestration: how to incorporate daily platform signals vs weekly incrementality results vs quarterly MMM output - Creative as a portfolio: how to weight hit rate & EV alongside typical ROAS metrics - Constraint-aware bidding: how to bake margin, inventory, compliance, and brand safety costs into the decision function
5 Comments -
Amirteymour Moazami
Cloaked has raised a $375M Series B led by General Catalyst with participation from HUMAN CAPITAL and Lux Capital. I am fortunate to have the opportunity to work alongside Arjun Bhatnagar, Abhijay Bhatnagar, and the team at Cloaked to fight against surveillance, spam, & identity theft.
1 Comment -
Irakli Kashibadze
Sustained ~252–291 GiB/s of HBM memory throughput on H100s under decode load — essentially hitting the hardware roofline. This matters because HBM throughput, not FLOPs, is the real bottleneck in LLM inference. By keeping memory nearly fully saturated, I’ve unlocked far higher efficiency and throughput than standard engines. The result: 0.9–1.36M tokens/sec with ~0.1 ms first-token latency #AI #LLM #GPURouter #H100 #Inference #CostEfficiency #Innovation #DeepLearning #AIInfrastructure #HighPerformanceComputing NVIDIA AMD OpenAI Google Shilpa Kolhatkar Keith Strier a16z speedrun
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content