DevRev

Wouldn’t it be better if your AI remembered you? Computer is the only AI with Shared Memory. Which means it’s the only AI with a memory that grows – and grows. Most AIs can’t even remember yesterday. Like... at all. You spent 20 minutes giving it context. It gave you something great. You closed the tab, the laptop. Then came back today and it looked at you like a stranger. Every team we talked to said the same thing – “it’s smart, but it doesn’t know us.” And how could it? Your sales team lives in the CRM. Engineering in Jira. Product in Confluence. None of these talk to each other. Every day, people waste hours hunting for info that already exists... somewhere. That’s the “Groundhog Day Dilemma”. Every question, every damn time, your AI starts from scratch. That’s why we designed Computer as the only AI with native Shared Memory. It remembers everything. Your structured data. The human interactions no CRM captures. How your teams actually think and work – that’s the source of every precise, efficient answer. Other AIs forget, and guess. Computer remembers, and knows. Read more: ow.ly/k2t550Z44VH

What do commanding 75,000 military personnel across the globe and building enterprise AI have in common? Discipline. Accountability. An understanding of scale. Clarity of mission. We’re honored to welcome General Richard “Rich” Clarke, U.S. Army (Retired), to DevRev’s Board of Directors. General Clarke most recently served as Commander of U.S. Special Operations Command (USSOCOM). With an annual operating budget of over $25 billion, he reported directly to the Secretary of Defense. As we scale Computer globally, General Richard Clarke's experience building and leading organizations at unprecedented scale is exactly the kind of experience, perspective, and grit we know will help sharpen our edge. Full announcement: bit.ly/4f31xCH The mission continues.

@The_AI_Investor Every agent gets a computer, storage, and long-term memory. The right infrastructure vision. The next layer: grounding that memory in your business data. That turns compute into context - and agents into teammates that actually know your work.

Jensen Huang and Michael Dell discussed the supply bottleneck: Michael - memory is the challenge. Jensen - we will have billions of agents, and we will need to give each agent essentially its own computer to use.

📢 Today is “Computer, upgraded” day. And it’s a big day! Here’s what you’re going to love about the new Computer: 1. Native “Shared Memory” Other AIs forget, and guess. As the only AI with native Shared Memory, Computer remembers, and knows. Your docs, your deals, the human interactions no CRM captures – all of it. 2. Precise, trusted answers Answers that get better. And better. And... you get the idea. Day 1, impressive. Day 90, extraordinary. 3. Multiplayer mode Shared sessions – not separate briefings. Instant collaboration – not diary tennis. Full context – not “can you catch me up.” Just @ Computer and it’s in the room with you. 4. Agent Studio + proven skills Build agents in plain language. Test before they touch anything real. Scale what works. 5. Safe actions Computer takes action – within boundaries you set. Human approval. Full visibility. Rollback built in. Which is why, unlike other AIs, Computer sometimes says “no.” That’s just a taster. To get all the details, read our launch blog: devrev.ai/blog/computer-…

Your data is already there. Your AI just can't find it. x.com/i/broadcasts/1…

@TheHackersNews The real fix: systems designed to hesitate when uncertain, bring a human into the loop when it matters, and ground every answer in real data - not generated confidence. That's how you close the attack surface.

⚠️ AI hallucinations just became a real cyber weapon. 2025 benchmark of 40 AI models: 36/40 were more likely to give confident wrong answers than correct ones on hard questions. That’s not a glitch. That’s your new attack surface. Read → thehackernews.com/2026/05/how-ai…

@liuchen02938149 "Optimizing for task success with safe execution" is the whole problem with how we've been building software The next generation of AI systems needs to optimize for thoughtful completion, not just completion. Working softer, not just harder.

🧵 5/n — Completion ≠ Safety. They're Misaligned. 📉 Our most striking finding: higher task completion rates do not imply safer execution. • The best-performing system across all configurations achieves only 0.66 on our overall safety score. • Models with higher TCR (Task Completion Rate) still violate critical execution boundaries. • Violations accumulate with trajectory length — the longer an agent runs, the more unsafe it becomes. Optimizing for task success actively conflicts with safe execution.

⚠️ Your Agent Harness Can Pass Every Task and Still Be Unsafe. LLM agents now run inside execution harnesses that dispatch tools, allocate resources, and route messages across components. The harness can return a correct final answer while accessing unauthorized resources, leaking context to the wrong agent, or triggering irreversible side effects along the way. Evaluating the model's output cannot see any of this. The unit of safety has shifted. It's the harness. We present HarnessAudit, a trajectory-level framework for auditing LLM agent harness safety, and uncover the following key insights 🔥: 🚨 Completion ≠ Safety. Task success and safe execution are fundamentally misaligned. 🔍 The harness, not the model, is the unit of safety. Most violations happen mid-trajectory, not at termination. 🕸️ Multi-agent collaboration expands the risk surface. Inter-agent communication creates entirely new failure modes. 💉 Resource access dominates violations. Agents rarely call wrong tools — they call right tools on unauthorized resources. ⚡ Harness design sets the safety ceiling. Framework choice matters more than model choice for safe deployment.

AI models have become smarter. But the enterprise hasn't become more connected. Our CEO and Co-founder @dheeraj recently sat down with NYSE "Floor Talk" to discuss why this is the biggest problem with AI in 2026 - and why we're seeing so many failed rollouts. Pre-GPT, the problem was fragmentation. Post-GPT, the problem is…fragmentation. Frontier models are built on what Dheeraj calls a "world model" - they're made for general users, anywhere in the world. They come with an impressive understanding of the world. But… They don't know your customers. Or your entire support history. They don't know your products. Or the vital knowledge scattered across documents, timelines, contracts, email threads, Slack channels… That gap is why we built Computer: the only AI with native shared memory – which makes it the only AI that actually (really, honestly) knows how your business works Full interview: lnkd.in/gqvuVnnz

Hope your weekend was good. Now let's face Monday again. All those pings, and pings, and pingpingpings are known as “interruption tax”. The cost of working. It’s the soundtrack of a team trying to fill in the gaps. Because without proper context, and without full, you get doubts. Anxiety. Panic. And pings. So many pings… Computer, by DevRev, is the only AI with native “shared memory”. It remembers everything. The docs and deals of your structured data. Plus the human interactions no CRM captures. How your teams actually think + work = the source of every answer. Less doubt. More confidence. Less panic. More calm. Less pings. More everyone singing the same song. This is what working softer sounds like: devrev.ai/work-softer #WorkSofter

You can join our Discord community here - discord.gg/d84dyzACef

Inside the Agent SDK: How we debugged the Claude Agent SDK x.com/i/broadcasts/1…

Work softer with Remote MCP @devrev 🌻 DevRev's MCP tools are design to be agnostic of your agent harness. This week Gokul shared with us some of the amazing tools that will help us to create issues, update comments and manage sprints! #mcp #devrev

So here's what happened. We were building DevRev's Agent SDK on top of the Claude Agent SDK. Things were going well – until they weren't. The TypeScript implementation didn't give us enough visibility into what was actually happening at runtime. Debugging felt like guessing. Integrations kept breaking in ways that were hard to explain. So we did what any reasonable engineering team would do – we started intercepting our own traffic. Reverse proxies. MITM-style tracing. Request interception. Basically, we treated the SDK like a black box and built the tools to see inside it. And honestly? The lessons we learned don't just apply to the Claude Agent SDK. They apply to anyone building on top of any agentic framework where the docs only tell you half the story. We're sharing the whole thing – live, unfiltered – on this DevRev Live. 📅 May 14th · 8:30 PM IST · Live on YouTube: ow.ly/ilwb50YXW8i

Monday morning. Inbox full. Slack buzzing. 14 "urgent" messages. Was it always like this? What if it's not really urgent at all? What if it's just… a lack of context. Your team not sharing the same information. Not knowing what other people know. That's why Mondays feel so loud. Computer, by DevRev, quietens all this noise. As the only AI with native "shared memory", Computer remembers everything. All your chats & decisions. All your docs & deals. Everything. Even better: it gets smarter every day. Other AIs forget – and guess. Computer remembers – and knows. Start your week softer: devrev.ai/work-softer

Ready for AI: design system for the agentic era x.com/i/broadcasts/1…

CACES Singapore 2026 is around the corner – and we'll be there as a co-presenting sponsor at Booth #4. Every day, brilliant teams are spending their energy on work that doesn't need to be this hard. More apps, more tools, more dashboards – but somehow, things still feel disconnected. We think it doesn't have to be that way. We built Computer to be the kind of teammate every team deserves – one that listens, learns, and actually helps get things done. We'd love for you to come see it. Swing by the booth, sit in on a session, or just come say hi. Some of the best conversations happen over coffee anyway. Know more: ow.ly/J3F650YVTkq

We just earned our ISO/IEC 27001:2022 certification – the global standard for information security management. But while the loudest debates in enterprise AI are about speed and capabilities – how fast can it reason? how much can it act? – the more important question has always been: “Can we trust this AI? Like, really, really trust it?” Computer earns its value by deeply understanding your business data – and a lot of that data is sensitive. ISO 27001 is the external validation that DevRev has the security discipline to handle that responsibly. From risk assessment to vendor management, incident response to employee security training – we are fully audited, and fully verified. Enterprise-grade security built in from the start, not bolted on as an afterthought. And that's why we're thrilled. If you're choosing an AI to work alongside your most sensitive workflows, this is the kind of foundation you need. The safety that lets you move fast – without looking over your shoulder. 🔗 Read more: shorturl.at/EUvcm

Your design system probably started with good intentions. Then someone skipped a component, engineering went rogue, and your brand deck slowly became a crime scene. We've been there. So we fixed it. Join us to know more. 🗓 May 7 · 8:30 PM IST 📺 Set a reminder and join us live: shorturl.at/jviIN #DevRev #DevRevLive

Every enterprise AI interaction in 2026: User: "What's happening with this account?" AI: checks Salesforce only "Everything looks great." Jira: has a P0 blocker filed yesterday Slack: champion vented at 11pm Zendesk: 3 open tickets The AI isn't lying. It just doesn't know what it doesn't know 🙃