OLX

Last week, we launched Otodim – a dedicated real estate platform for Ukraine. 🏠🇺🇦 This one matters to me personally. Ukraine is not a market we're managing from a distance. We have 150+ people on the ground, we're one of the top 100 taxpayers in the country, and we've invested well over $10 million in direct support since 2022. And the market is responding: ➡️ 92% of pre-war active users back on the platform ➡️ Active listers recovered to 99% of pre-war levels ➡️ Buyer engagement on apartment listings up 12% year-on-year That's not a slow recovery – that's a market that is vibrant and intrepid. Otodim is the natural next step. A dedicated experience for property seekers and realtors – built on OLX Ukraine's full listings inventory, AI-powered, mobile-first, and designed to simplify a genuinely complex housing market. This is continued investment in a market we believe in, backed by real growth and real potential. Real Estate is one of OLX's three core growth categories. Revenue up 26% in H1 FY26. We run dedicated property platforms across Poland, Portugal, Romania and South Africa – and Ukraine now joins that group. Same playbook across all our markets: deep local investment, category focus, product built on real user behaviour. We believe in this market. We're building for it. Today and in the long term. #OLX #Otodim #Ukraine #RealEstate #Marketplaces

Most companies lie about tech debt time. Photoroom actually put it on the calendar. One underrated part of Photoroom’s engineering culture: Fridays are dedicated to platform work. Not “we care about tech debt” in the abstract. Not “engineers can spend 20% of their time on internal improvements”, which usually means: “Sure, you can do it after the roadmap, the bugs, the urgent requests, the surprise urgent requests, and that one thing someone just remembered was critical for yesterday.” Actual time. Every Friday, engineering gets space to work on engineering topics. When I joined Photoroom, this was explained very clearly from the beginning. And honestly, I was skeptical. I’ve seen this promise before. Most companies say they want engineers to keep the codebase healthy. They say internal tooling matters. They say platform work is important. Then the product roadmap gets louder, someone says “just this once”, and suddenly your dedicated time becomes an endless project in Linear. What makes it work here is that the product team actually respects the arrangement. Friday is not treated as leftover product capacity. It’s engineering time. And that changes the dynamic completely. But this is not just “free time for engineers to do whatever”. On the engineering side, it comes with responsibility. We define the priorities. We pick the problems that matter. We keep the foundations in good shape. Better tooling. Cleaner architecture. Faster builds. Less flaky tests. Healthier dependencies. Better developer experience. The kind of work that rarely gets a launch post, but absolutely shows up in how fast and safely the team can move six months later. And the best part: every engineer is encouraged to own platform projects. Not just “the platform team”. Not just “the senior-most person who remembers why this folder exists”. Everyone. That’s how you build ownership beyond features. Because a strong engineering culture is not only about how fast you ship. It’s also about whether you leave the system better than you found it. #photoroom #engineering #softwareengineering #platformengineering #frontend #developerculture

Product is about solving hard customer problems. Monetisation is about responding to these through the offering and purchase experience. There's no way to drive P&L impacts without understanding sales. Huge appreciation to Dr. Dominikus Kirchhoff for an amazing session today! Big thanks to Marcin Kawecki for the invitation.

🖥️ Last Tuesday, I found myself debugging code alongside other CEOs on a video call. Not something I ever expected to be doing. Fabricio Bloisi had all the CEOs and Senior Leaders from Prosus build AI agents. Not watch. Build. I'll be honest: I hesitated and felt uncomfortable at first. I'm a guy with a commercial and finance background who transitioned from CFO to CEO over time. I don't code. Never thought I would. That's what we hire developers and engineers for, right? 🤷 But then it clicked: if we're building an AI-first company, the CEO can't stay on the sidelines. So I built three agents that day. Two focused on the automotive market potential in Europe. The third? An org design agent I'm still working on. Here's what surprised me: - No briefing. No waiting. No iteration loops. I built it, tested it, and moved on — same day. ⚙️ - I felt closer to the problem, which meant closer to the customer — and faster than I was two weeks ago. 🚀 - It was actually fun! 😄 That's also because ‘my professor’ Vitor Castro helped me navigate through countless unexpected questions that came up when I started diving deeper into a subject that was totally new to me. Turns out you don't need a technical background to solve problems with AI. You just need to start and use it consistently. The org design agent isn't finished yet, but I'm curious about where it leads. It's complex, so I teamed up with our Chief Data Officer, Andreas Merentitis , and our CHRO, Marta Jakowluk . With AI already handling repetitive work, the real opportunity lies in redesigning roles around what humans do best — judgment, creativity, critical thinking, and customer connection. The best part? This conversation is already happening across the whole company at OLX — not just in Product, Tech, and Data, but everywhere. That's the opportunity I'm thinking about, and it all feels so much more possible now. #OLX #AI #Prosus #Leadership #Marketplaces

𝗧𝗵𝗲 𝗔𝗜 𝗦𝗲𝗿𝘃𝗶𝗰𝗲 𝗥𝗲𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝗶𝘀 𝗛𝗲𝗿𝗲: 𝗠𝗲𝗲𝘁 𝗢𝗽𝗲𝗻𝗔𝗜’𝘀 "𝗗𝗲𝗽𝗹𝗼𝘆𝗖𝗼" 🚀 OpenAI is no longer just building the world's most famous models—it’s now moving into your office to help you run them. The launch of the OpenAI Deployment Company (DeployCo) marks a massive shift as the AI giant officially enters the services industry. Here is what you need to know about this $4 billion venture and why it’s shaking up the global IT landscape: 𝟭. 𝗧𝗵𝗲 $𝟰 𝗕𝗶𝗹𝗹𝗶𝗼𝗻 "𝗦𝘁𝗿𝗶𝗸𝗲 𝗧𝗲𝗮𝗺" DeployCo isn’t just a small experimental wing. It has launched with over $4 billion in investment from heavyweights like Goldman Sachs and SoftBank. While OpenAI maintains majority ownership and control, the company is built to be a powerhouse of implementation. 𝟮. 𝗘𝗻𝘁𝗲𝗿 𝘁𝗵𝗲 "𝗙𝗼𝗿𝘄𝗮𝗿𝗱 𝗗𝗲𝗽𝗹𝗼𝘆𝗲𝗱 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿" (𝗙𝗗𝗘) Forget standard tech support. OpenAI is deploying a specialized workforce of 150 Forward Deployed Engineers and Deployment Specialists. Their mission? To work side-by-side with businesses to understand their specific needs and handle the heavy lifting of integrating AI tools into existing internal networks. 𝟯. 𝗧𝗵𝗲 𝗘𝗻𝗱 𝗼𝗳 𝗧𝗿𝗮𝗱𝗶𝘁𝗶𝗼𝗻𝗮𝗹 𝗜𝗧? 𝗡𝗼𝘁 𝗤𝘂𝗶𝘁𝗲. The move has sparked talk of a "SaaSpocalypse" for Indian IT giants like TCS, Infosys, and Wipro. On the surface, DeployCo looks like a direct competitor, allowing OpenAI to bypass traditional service providers. However, many experts—including Box CEO Aaron Levie—believe this is actually a "blessing in disguise". Levie argues that the shift to AI agents is a transformation on a scale "far greater" than the move to digital or cloud. 𝟰. 𝗪𝗵𝘆 𝗗𝗼𝗺𝗮𝗶𝗻 𝗘𝘅𝗽𝗲𝗿𝘁𝗶𝘀𝗲 𝗦𝘁𝗶𝗹𝗹 𝗪𝗶𝗻𝘀 While OpenAI owns the core technology, established IT firms have spent decades mastering the customization of technology for complex client environments. The market is proving to be massive; OpenAI already has over one million businesses using its products and APIs, creating an "insane amount" of technical and domain-specific work that requires more hands than any one company can provide. In fact, firms like Infosys and TCS are already landing deals to collaborate with AI leaders like OpenAI and Anthropic to drive this change. The Bottom Line: We are witnessing a new era of "Agentic AI." OpenAI isn't just handing over the keys to the car anymore; they are building the infrastructure to make sure every industry has its own custom-tuned engine. What do you think? Will DeployCo dominate the service market, or will traditional IT giants remain the kings of customization? Let’s discuss in the comments! 👇 #OpenAI #DeployCo #GenerativeAI #ITServices #DigitalTransformation #TechNews #FutureOfWork

(LONG TEXT ALERT, but I think it's worth it 🙃) 3 situations I witnessed (or lived) that suddenly made total sense to me (just today 👀 ): 1. A designer who used to hire a frontend developer to build websites. Now she does almost everything herself with vibe coding and only asks the developer for minor adjustments. (Important note: also many bug fixes. That part didn't change.) 2. A Technical PM who was quietly freaking out because of a lack of technical fluency. Now, with an AI coding assistant, he has the intelligence of a senior developer and architect in his pocket (and he's absolutely killing it). 3. A Product Manager (hi, that's me 👋) who has been trying to stop depending on data teams for every single SQL query. AI is opening possibilities we couldn't have predicted even a few months ago. It's moving fast (uncomfortably fast). And while some brilliant colleagues seem to be flying with AI adoption, I was stuck trying to understand, conceptually, what this all means for my role. Then today our CPO Michael Bullion shared something that finally gave it a name. 🗣️ Two quotes that landed: - "The rise of agentic AI represents a foundational change in how we'll work" — Fabricio Bloisi (Prosus) - "AI transformation isn't about sprinkling AI everywhere. It's about fundamentally changing how we work" — Christian Gisy (OLX) 💪 And the three PM muscles that matter most in this era: - Product Judgment: making good decisions with incomplete data, balancing instinct with evidence - Prototype Building: building to learn, not to ship - Strategic Sense:deciding what not to build in a world of infinite possibilities But the thing that really stuck with me? We no longer have designers designing, PMs defining, and engineers building. We have fluid roles. FLUID ROLES. 🤯 Obvious to many, I know. But it really just clicked for me TODAY. And somehow putting a name on it makes the whole thing feel less chaotic and more like an invitation. So here I am on a Monday evening. Lots to digest. But honestly? I'm kind of here for it. 🤷♀️

TL;DR: Shit just got serious. I'm 40. Let's skip presents this year. Hire me as a Facilitator or UX Researcher instead. I know, I know. I don't look a day over 27. A year ago I had no idea my life would look like this. That I'd be laid off. That I'd be figuring it out again from scratch. That this would be the year of both anxiety and excitement. Big numbers in my life seem to come with turns. Ten years ago, just after I turned 30, my dad died suddenly. I'd been thinking about moving into UX for a couple of years, but kept telling myself I was too old to change my career path, which sounds pretty ridiculous now. After he died, something snapped. I went to one of the companies I was teaching English at and (long story short) asked if they'd hire me as a UX Researcher and it was the best decision in my life. Turns out UX Research and facilitation is what I really love doing. And I don't want to brag, but I'm pretty good at it. Now I'm 40. Freelancing. Figuring it out again and loving it.The pattern holds. So again, no need to buy me stuff. Just ping me and I'll turn your users, data, and internal knowledge into decisions you won't regret. (original photo taken by the one and only Gosia Jakubiak)

🚀AutoGPT is live for OTOMOTO , Standvirtual and Autovit.ro ! We've launched our conversational AI assistant for car buyers in Poland, Portugal and Romania – built in partnership with OpenAI and accessible directly through ChatGPT. This is the second agentic AI assistant we've launched this year, following CompassGPT for property search. Up next: France. 🇵🇱🇵🇹🇷🇴🇫🇷 It is trained on proprietary transaction data from millions of verified OLX Motors platform deals across all of our geographies, giving consumers the real price, vehicle history, and everything they need to know to make the best choice when buying a car. This is market intelligence that generic LLMs don't have. Real data from real transactions. 🚗📊⚡ AutoGPT completes the two-sided vision: AutoIQ already helps dealers optimise operations. Now buyers get AI-powered conversational search. Both sides of the marketplace are running on agentic AI. 🤖⚡ We've invested over $200M in AI since 2018, and the momentum is accelerating. We have 75 active use cases, of which 10 are agentic applications. OLXers have built more than 3000 agents to help us with productivity and better services to our customers and consumers. This is what building the future of automotive commerce at scale looks like. 📈🚀 How does it actually work? What makes it different? Where is this heading? Watch the video 👇 #OLX #OTOMOTO #AutoGPT #AgenticAI #Automotive #Marketplaces

Just saw what the team shared from Braze City x City in London. BrazeAI Operator™ might be the shift we've been waiting for 🤓 . Global builds the agents and frameworks—local teams deploy with market context. Speed without losing control. Hope the languages support works well 😉 . --> Getting ready to start building. https://lnkd.in/eKiuBHr4 #BrazeAI #CRM #MarketingAutomation #AI