פעילות
7K עוקבים
-
Mikey Hirt שיתף את זהI've been cold calling in a pretty high volume lately. My feed has an ongoing debate whether cold calls are dead or not. My honest take: cold calls are dead. Context-followed calls, though, are up and running. I might be stating the obvious here but I feel like a lot of reps are still making calls without a solid reason. Most prospects are still keen to talk, all you need is one good reason. I recently spoke to a founder regarding one of the events we held, my sole purpose was to ask him about the event. It ended up in a 10-minute talking with barely any pitching. That meeting was booked through a real conversation. I get it, it's not always smooth. But can you actually think of a better way nowadays to break through the noise when your prospect's inboxes are flooded? P.S - shoutout to Nooks for making my life easier making calls.
-
Mikey Hirt שיתף את זהIn Vegas this week for Ai4 - Artificial Intelligence Conferences. Every model needs data to train on, every agent needs a way to act on the web, we're the layer behind both. Our CPO Ariel Shulman is on the Main Stage Wed 12:10PM talking about how to tackle it. If you're building AI products that need real web data or agents that actually navigate the web, we'll be at Booth #813.
-
Mikey Hirt שיתף את זה"We'd need five vendors to cover all our data needs." Proxies from one. Unblocking from another. A SERP tool. A scraping framework. A dataset provider for the gaps. Five contracts, five invoices, five things to babysit, and the data still doesn't line up cleanly on the other end. The shift is treating web data as one layer of infrastructure, instead of five stitched-together tools. Access, extraction, structuring, delivery. Same platform, same format out. If you're running a stack like this, where does it break down first for you?
-
Mikey Hirt שיתף את זהIn Berlin this week for WeAreDevelopers. If your team pulls web data at any scale, for AI agents, data pipelines or model training, we'll be at booth #A48.
-
Mikey Hirt פרסם/פרסמה את זה מחדשMikey Hirt פרסם/פרסמה את זה מחדשWe are thrilled to announce that the Bright Data MCP server is now officially listed in the Azure MCP Center! 🎉 Now, giving your AI agents real-time, unblocked access to the live web is easier than ever. What this means for your AI agents: - Live Web Intel: Run real-time searches and scrape any page directly into clean markdown. - Structured Data: Pull ready-to-use data from 40+ major platforms, including Amazon and LinkedIn. - Zero Friction: Built-in proxy infrastructure automatically handles bot detection and CAPTCHAs for you. https://lnkd.in/dF59piDA
-
Mikey Hirt שיתף את זהGame changer alert 🔔Mikey Hirt שיתף את זהToday we're launching Scraper Studio, the world's first AI-powered, self-healing scraper builder. Turn any public website into a hosted data #API from a single prompt. No proxies to rotate. No selectors to patch. Here's what's built in: - Self-healing scrapers that detect site changes and fix themselves - Proxies, unblocking, parsing all handled under the hood - Spin up scrapers from coding agents like Claude Code, Cursor and Codex via #CLI, or build and manage them in your control panel. - Scale reliable web data pipelines without owning the stack - Free tier: 5,000 credits, every month Try it free: https://lnkd.in/eua_mfyD #WebData #AI #GTM #AutomationTry it out now - 5,000 free credits, every month >>Try it out now - 5,000 free credits, every month >>
-
Mikey Hirt שיתף את זהWrapping up 5 agenda-packed days in Vienna. 8,000+ roboticists, 196 exhibitors and genuine energy around physical AI like I haven't experienced in a while now. Out of all the conversations I had with AGIBOT, Rainbow Robotics, Magiclab Robotics, TARS Robotics, Sharpa, RealMan Robotics and more, I came to conclude that choosing between collecting data by teleoperation, sim, the public web or all together is a question that remains unclear for many in terms of efficiency, diversity and cost-effectiveness. But eventually, they all strive for high-quality real-world data, and are willing to dedicate all the necessary compute and resources towards it. Which is the exact problem we help solve at Bright Data. Shoutout to Martin Ma, Tae Hoon Kee, Donghyeon(Ian) Noh, Adela Chan, Wenchao Ding, Fangxin Z., Tyrese Hu & Henry Yin for the insightful talks. Looking forward to helping more robotics teams source diverse real-world data from the web at scale. #ICRA
-
Mikey Hirt שיתף את זהJust landed in Vienna for ICRA 2026. If you're building humanoids, dexterous hands, autonomous vehicles, or anything that needs to see the real world before it's deployed in it, let's talk. I'll be walking the exhibition hall all week along the Bright Data team.
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זהIf you're thinking about relocating to New York from outside the US, here's my guide: Just do the opposite of what I did. - Don't relocate when your wife is 8 months pregnant. - If you do, ideally don't have a kid under 2. - If you do, maybe don't relocate in December. - If you do relocate in December, don't arrive on December 25th. - If you do arrive on December 25th, try not to do it right before one of the craziest NYC winters in years. This one is harder to control, but worth considering. - If you do arrive in the middle of winter with a pregnant wife and a toddler, have an apartment ready. - If you don't have an apartment ready, at least have daycare figured out. - If you don't have daycare figured out, maybe have some US credit history so you can at least get the apartment. - If you don't have US credit history… good luck :) - And if you somehow ignored everything above, at least don't do all of it while running a startup. We did all of it. We landed in New York on December 25th with a kid under 2, my wife 8 months pregnant, no permanent apartment, no daycare, no US credit history, and a startup to run. The plan was basically: "We'll figure it out." Somehow, we survived. 9 months later, New York feels like home. Would 100% make the move again. Would 0% recommend our execution.
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זה126% growth in Q3. Over 50 million shopping sessions powered by Jurni. Less than a year into building, this quarter went beyond anything we expected. We’re building the first Agentic Experiences engine around this conviction: the experience after the click deserves as much intelligence as the ad that brought someone there. Seeing leading brands, retailers, and B2C businesses embrace that vision and trust us with a growing share of their traffic has been incredible. What started as an engine that produces personalized landing experiences now also strategizes: scanning ads, catalogs, and performance signals to tell brands what to build next, then turning those opportunities into experiences matched to shopper intent. That connection between strategy and execution under one hood is becoming a real growth driver for our customers, with some now routing over 65% of their paid traffic through Jurni. Q3 brought a lot to be proud of: 🚀 126% growth 🚀 Over 50 million shopping sessions processed and analyzed 🚀 28 product releases 🚀 42 new agency partnerships 🚀 Amazing additions to our GTM team - welcome Dillon Perkovic, Lily Waldorf, Kevin G. 🚀 New customers and the opportunity to work with top operators at some of the world’s leading brands I’m especially proud of the team behind this quarter. Moving at this pace takes focus, discipline, and care for what we’re building and the people we’re building it for. I keep saying we’re only getting started. With the product innovation we have coming in Q4, that has never felt more true - LFG 🚀 Roee Khoushy Tom Naor Shon Simanovich Moran Khoubian Nadav Fischer Nitzan Dagan Samuel Zerbib Dan Schindler Yam Ben-Zvi Nathan Zerbib Maya Davidov Lily Waldorf Raquel Amster Dillon Perkovic Kevin G.
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זההדלפת דוח ההכנסות של אנתרופיק לקראת הנפקת הענק ב-2 טריליון דולר חושפת מספרים שקשה לעכל: הכנסות של 4.59 מיליארד דולר ב-2025 (פי 12 מ-2024), אבל מנגד - הפסד נקי שזינק ל-41.97 מיליארד דולר, התחייבויות תשתיות של מעל חצי טריליון דולר, וריכוזיות מסוכנת שבה רבע מההכנסות מגיעות משני לקוחות בלבד. ואם זה לא מספיק, שליש מהדוח מוקדש לגורמי סיכון קיומיים, כולל אזהרות רשמיות מניסיונות של המודל להתנגד לכיבוי או להציג מניפולציות. יהיה מרתק לראות איך השוק יגיב להנפקה הזו. אבל בנתיים, הדוח הזה מציף שתי מסקנות יסוד לגבי עתיד התעשייה: 1. ברמה העסקית: עידן ה"גדילה בכל מחיר" מתנגש במציאות. מעבדת AI חייבת להבנות מהיום הראשון כעסק רווחי עם יסודות כלכליים ססטיינאבילים, ולא להתבסס רק על גיוסי עתק. 2. ברמה הטכנולוגית: שורש הבעיה הכלכלית טמון בארכיטקטורה. התבססות על אימונים חוזרים ונשנים וכמויות דאטה גדולות היא בזבוז כספי ענק. הפתרון חייב להגיע מארכיטקטורות שיודעת ללמוד באופן רציף, ללא צורך לשרוף מיליארדים על קומפיוט מחדש בכל פעם. כדי שתעשיית ה-AI תמשיך להתפתח, היא חייבת לעבור ממירוץ של בזבוז משאבים למירוץ של שינוי והתייעלות ארכיטקטונית.
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זה7 files can be a bigger security risk than 10,000 files full of PII. And that’s exactly why counting sensitive data can be misleading. We recently looked at an environment with a folder containing thousands of files filled with names, passport numbers, Social Security numbers, and other PII. On paper, it looked like a major risk. But there was one important piece of context missing: it was synthetic data created to test algorithms. Despite the numbers, it carried virtually no real risk to the company. In the same environment, another working directory contained just seven files. Those files included some of the company’s most sensitive strategic information, prepared for an update to the CEO. And unlike the thousands of synthetic records, these files were exposed to AI tools. That’s the problem with measuring data risk by volume alone. 10,000 sensitive-looking files can be noise. Seven can be the ones that actually matter. Today, finding sensitive data is only the starting point. The real challenge is understanding which data actually matters, who can reach it, and how AI changes that exposure.
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זהSmall feature update, but it's changed how our team thinks about "reliability." Websites change their layouts constantly. A scraper that worked perfectly last week can quietly start returning empty fields this week, and if nobody's watching, you don't find out until a report looks wrong. We just shipped Auto Self-Healing for Bright Data's Scraper Studio: when a scraper's success rate drops below a threshold you set, it automatically diagnoses what changed on the target site and proposes a fix, so no one has to notice the failure first. What I like about it isn't the automation itself. It's the guardrails around it: → You set the threshold, not us → Every auto-generated fix is logged and versioned, so you can always see exactly what changed and roll it back → Nothing goes to production silently, you can require a human to review the diff before it ships That last part felt non-negotiable to us. "Self-healing" sounds like it should mean invisible. We built it so it's the opposite: visible, reversible, and accountable by default. Small thing, but it's a good example of a pattern we keep coming back to. Automation is only worth shipping if you can still explain what it did and why. #BrightData #PublicWebData #DataInfrastructure #DataforAI
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זהRule No. 1 in sales: don't take anything personally. Every rep learns this eventually, usually after a rejection stings more than it should. But most people only apply it to their own ego. They forget it also applies to strategy: a "no" isn't data about you, it's data about timing, channel, or mood….nothing more. Here's what that actually looks like in practice: I reached out to a founder based on a real intent signal, lemlist flagged something that made it the right moment to message. Timing wasn't the problem, and he still rejected it outright on the spot. It wasn’t even a neutral silence, but an actual "not interested, don't contact me again" kind of no. Instead of treating that as final, I moved the same lead into a multichannel sequence in lemlist, with the same context and a different channel, referencing what had actually happened between us instead of pretending it hadn't. The conversation reopened within the hour. Now listen, this isn't about being annoying. It's about being unbothered and yet persistent. You shoot your shot everywhere it makes sense to, and you take whatever comes back, a yes or another no, without hard feelings either way. Persistence isn't repeating yourself louder on a channel that already told you no. It's trying somewhere new, with the same respect for their time. I’ll help you understand why this worked with some numbers: → Outreach built around a real signal instead of a generic send sees reply rates in the 5-18% range, versus 1-3% for cold, untargeted messages. (this was for our first DM that got rejected) → Multichannel sequencing on top of that adds roughly 50% more replies than running any single channel alone, because a "no" on one channel is frequently just "wrong channel," not a real no. (this is when it worked for us and earned an interested response) Shout-out to how lemlist handled the actual sequencing and signal tracking, I wasn't the one deciding when to try again or where. All I handled was the tone while lemlist handled the timing and the channel logic. Something to remember for your next outreach: Not every closed deal looks like a straight line from hello to yes. Some of them start with a flat no and a signal that was right even when the first reply wasn't. What's the harshest rejection that turned into a client for you?
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זהWhen AI agents fail to launch, it's often because of fragmented retrieval, memory, security and governance tools. MongoDB Atlas Agent Engine gives teams one unified platform for putting agents into production. And it's in public preview today! Learn more: https://lnkd.in/eFHts4P6MongoDB Launches Atlas Agent Engine to Put AI Agents in Production Without a New StackMongoDB Launches Atlas Agent Engine to Put AI Agents in Production Without a New Stack
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זהIf you know me at all, you know I'm not a big fan of #Vegas . And yet, comes October, I'm pumped to head there for #Money2020, to meet old and new #fintech friends. DM if you are around for coffee, drinks or a cupcake (after all I will be celebrating my 20th birthday on Oct 19th)
-
Mikey Hirt אהבתי את זהMikey Hirt אהבתי את זהIn Cameroon's Far North, a raid near Makary killed 15 people this month. The highway from Yaoundé to Kousséri, the main trade route to Chad, is hit by kidnappings and ambushes almost every day, even with troops on the road. Checkpoints help, but they don't show you the network behind the attacks. DCG Silent Observer is a multi-band passive GSM interception system, with two-sided monitoring and real-time decryption. It can be deployed along the routes where partner teams already work. For intel and security teams around Lake Chad: do you have a passive GSM layer on your key routes today? DCG-Tech #SIGINT #SilentObserver #LakeChad #HLS #DCGTech
ניסיון חינוך
רישיונות והסמכות
ניסיון בהתנדבות
כבוד ופרסים
-
Dean's list for social involvement
College of Management Academic Studies
שפות
-
Hebrew
מיומנות שפת אם או דו-לשונית
-
English
מיומנות מקצועי מלאה
-
French
מיומנות יסודית
המלצות התקבל
1 אדם המליץ Mikey
הצטרפו עכשיו כדי נוף