# Product Builder > Cloudtalk ยท Czechia (Remote) ยท โ€” ยท Posted 2026-09-14 **Workplace:** remote **Department:** Product ## Description ### Product builder ๐ŸŒ Global SaaS | $28M Series B Investment ๐Ÿ“ Remote in Europe ### The Mission Most product ideas die in the gap between "someone should look at this" and "a squad has capacity next cycle." As a Product Builder, you close that gap by being the whole loop yourself: you find the problem, talk to the users, build the thing, ship it, and read the result. You own the path from problem to outcome. Not just the spec. Not just the ticket. The whole thing. This is not a Product Manager who writes some code, and it is not an engineer who sits in on discovery calls. It's a single role that holds discovery and delivery in the same head, and uses AI to make that possible at a speed a three-person trio can't match on exploratory work. You'll work in the messy, high-ambiguity spaces around our Core Platform Product, our AI Voice Agents, Calling Apps and our product-led surfaces - the friction nobody owns, the repeated support question, the idea that's too early for a squad to commit a cycle to and too valuable to drop. Our empowered squads own our committed roadmap and they're good at it. You own the things that haven't earned a squad yet. A good week for you isn't "I shipped a lot." It's "we now know something we didn't know on Monday, and there's working software proving it." ### ** Key Responsibilities** ### 1\. Own the problem before anyone has written it down - Find the friction. You go looking - in support tickets, in churn calls, in the questions that keep landing in #product, in the gaps between two teams' surfaces. You don't wait for a brief. - Sharpen it into a hypothesis. Vague frustration becomes a falsifiable statement with a metric attached, fast enough that the sharpening doesn't become the project. - Talk to users yourself. Continuous discovery is not a phase you schedule. You're in contact with real customers weekly, and you can run an interview that surfaces behaviour rather than feature requests. - Run several bets in parallel. You explore three or four opportunities at once, then converge on the one or two worth real investment. ### 2\. Build and ship it - End-to-end delivery. Prototype, build, test, ship to production, instrument. You work across the stack and you make your own design calls - there is no PM or Designer assigned to you, though both are a Slack message away and you use them well. - Ship small and early. You'd rather have something imperfect in front of five real customers this week than something complete in front of nobody next month. - Hold the quality line. Shipping fast is not shipping sloppy. You write tests, you think about failure modes, and you know when a prototype must not be allowed to become production. - Build AI-native, not AI-bolted-on. LLM-driven flows, agentic workflows and orchestration - you understand the models well enough to design around what they actually do, including what they get wrong. ### 3\. Read the result and decide - Instrument before you ship. You define what would prove you wrong before the code goes out. - Kill your own ideas. You can look at a two-week build, see it didn't move the number, and shut it down without needing anyone's permission or a retro to justify it. This is the hardest part of the job and the one we'll dig into hardest in interviews. - Hand off deliberately. When something works and needs to scale, you write it up, transfer it to the owning squad, and walk away. You do not accumulate a personal estate of unowned production services. ### 4\. Work AI-first, and make it count - AI as a multiplier, not the point. You use agentic workflows to compress the loop from idea to shipped. You can explain which tool for which job and why, and you know when the agent is confidently wrong. - Systems, not prompts. Repeatable workflows for research synthesis, prototyping, and experiment readouts - not one-off chats. Validation designed in. - Raise the floor. The patterns you find get written down and shared. Your leverage should show up in other people's throughput, not just your own. ### 5\. Stay connected while working alone - Visible by default. You work with high autonomy, which means you over-communicate. What you're exploring, what you learned, what you killed, all in the open. - Commercially honest. You can explain what a bet is worth to the business in ARR, retention, or cost terms, and you can say "this isn't worth building" out loud. - A good citizen of the roadmap. You don't parachute into a squad's surface without talking to them. Discovery you run in their domain gets handed to them.| The Ideal Profile - The Full-Loop Operator. You have personally taken something from "I noticed a thing" to "customers are using it in production" - and you can name the metric that moved. - The AI-Fluent Builder. You re-engineer how work happens with AI (Level 4 on CloudTalk's AI fluency ladder). You ship things to production that would previously have needed a whole trio. You'll be asked for receipts. - The Comfortable Killer. You've shut down your own work. You can talk about a bet that failed, what it cost, what it taught you, and how fast you called it. - The Discovery Practitioner. Continuous discovery in the Teresa Torres sense is how you work, not a framework you've read about. You've run enough interviews to know how badly a leading question can poison a week of work. - The Ambiguity Native. Startup, founding, or scale-up background. You've worked without a spec, without a designer, and without anyone telling you what success looks like. You prefer it. - The B2B SaaS Realist. You understand a sales-assisted and PLG motion, an ARR-driven model, and the difference between a feature a customer asks for and a problem worth solving. **Nice to have**: voice, telephony, or real-time systems experience. Founding or solo-technical-founder background. A public trail of things you've built. ### ** Success Metrics (First 90 Days)** - Shipped, not planned. Two or three validated experiments live in front of real customers, each with an instrumented result and a written call: scale, iterate, or kill. - At least one kill. You've shut something down and explained why. A Product Builder with a 100% hit rate in their first quarter wasn't taking real bets. - A hypothesis backlog. A visible, prioritised set of opportunities in your space that didn't exist before you arrived, sourced from your own customer contact. - One clean handoff. Something you built has been transferred to an owning squad with documentation, or deliberately retired. Nothing is orphaned on you. - **Loop time. You can state your median time from hypothesis to customer-facing test, and it's measured in days.** ### **About CloudTalk** - CloudTalk is THE global AI-powered business calling platform, trusted by 30,000+ professionals at companies like Nokia, Revolut, Glovo, and Uber. - We're building the world's most capable AI business calling software available in 160+ countries - helping businesses have more and better calls every day. Backed by $35.3M in funding from leading investors, including KPN Ventures, Point Nine, and HenQ, we're investing in making AI native to every conversation. - We're a ~200-person team that values growth, innovation, teamwork, customer focus, and human connection. We move fast, build for the future, and care deeply about creating products - and a culture - that people genuinely love. ## Benefits Why youโ€™ll love being a CloudTalker: - Unlimited PTO that people actually take. Fully remote with four European hubs. A coach you can book, a career path in writing, and equity when you deliver. Here's the full package: - Get better, faster - ๐Ÿ’ฌ A coach you can actually book. Recurring 1-on-1 sessions with our resident coach. A real conversation about your career, on company time. - ๐Ÿ› ๏ธ You grow by shipping, and we build around that. Roughly 70% of your development comes from hands-on work on real problems. We back the rest with coaching and formal training, following the 70/20/10 model. - ๐Ÿชœ Promotions without guesswork. Our job-leveling framework is transparent: every level and its expectations, in writing. You'll always know what the next step requires and where you stand. - ๐Ÿ“ˆ A path to equity. In R&D, equity is earned, and the bar is public: deliver exceptional results and impact, and you get your stake. When the company wins, you win with it. - Work the way you work best - ๐ŸŒด Unlimited PTO, and we mean it. Taking 30 days a year is normal here. Deliver beyond expectations, and you've bought yourself room for more. - ๐ŸŒ Fully remote, offices optional. R&D works remotely by default. When you want a desk, a whiteboard, or other humans, our hubs in Bratislava, Brno, Prague, and Barcelona are open to you. No office-days quota. - โฐ Flexible hours, built around your team. Start early or shift late as life needs it. Be there for meetings and your team's core overlap; how you arrange the rest of your day is yours. - ๐Ÿ’ป Gear that doesn't fight you. A MacBook for work, plus a home-office allowance to build a setup you like sitting at. - Rest like you mean it - ๐Ÿ”‹ Recharge Fridays: 4 extra all-company days off a year. Once a quarter, everyone is off at the same time. You come back to zero pings, not a doubled backlog. That's the point. - ๐Ÿƒ Fitness allowance. Gym, climbing, yoga, whatever keeps you moving. You choose, we pay. - People worth staying for - ๐Ÿ’ธ Up to โ‚ฌ5,000 per referral. You know great engineers. When we hire one of them, you get paid for it. - โœˆ๏ธ Two offsites a year worth packing a bag for. The all-company offsite (last year: [Tuscany](https://www.youtube.com/watch?v=RJYfbHH9Fak) ๐Ÿ‡ฎ๐Ÿ‡น) plus a dedicated R&D offsite, both somewhere remote enough that people actually talk. - โšก Hackathons twice a year. Two days to ship the idea you keep bringing up at lunch. - ๐Ÿ‘• Fresh CloudTalk swag every year, with a points system to earn more. Apply now and be part of our journey to $100M ARR! By submitting your application, you acknowledge that Cloudtalk will process your personal data based on Cloudtalk's legitimate interest in finding the most suitable candidate and a pre-contractual relationship with you. For more information on how Cloudtalk processes your personal data, please refer to our [Privacy Notice](https://www.cloudtalk.io/privacy-notice/). ## Apply [Apply at Cloudtalk](https://apply.workable.com/cloudtalk/j/8F45B904A7/apply) --- Powered by [Workable](https://www.workable.com)