Senior Applied AI Engineer - EU based

## About the role Help us build the future of conversational-based sales Weply is a leading startup in conversational-based qualified lead generation. We currently help more than 2000 companies in Europe grow faster, by converting relevant web traffic to qualified leads through a chat solution. We have human experts man the chats on behalf of the customers and use AI under-the-hood to supercharge the experts. We also help companies provide excellent support in non-leadable conversations. You will help leverage the unique dataset that Weply has to build the future of conversational sales conversations, utilizing the best of AI and humans in collaboration, to create a unique value proposition for customers and an excellent experience for end-users. ## What You'll Do - Own the end-to-end lifecycle of our AI systems. Working with our CTO, data scientists, and product team, you'll make pragmatic tradeoffs, ship iteratively, and continuously improve what users experience. - Design and build AI systems that work in the real world - Lead the architecture, design, and implementation of LLM-powered workflows: RAG pipelines, multi-step agents, retrieval systems, tool calling, and fallbacks - Make decisions about when to use LLMs and when not to, and defend those calls - Own reliability, latency, and cost tradeoffs across the entire stack - Handle real-world failure modes: hallucinations, edge cases, degraded responses, and the unexpected - Build for humans and at scale - Design systems that integrate deeply into our product and work seamlessly with human chat agents - Implement logging, tracing, and debugging for AI behavior - Define quality standards and guardrails together with our data science team - Build evaluation frameworks (offline and online) to continuously improve what users experience ## What We're Looking For ### Must-have - Shipped and operated LLM-powered systems in production. This is a non-negotiable. - +3 years of strong backend engineering skills (Python or similar) - Hands-on experience with RAG systems beyond basic demos, LLM APIs (e.g. OpenAI, Anthropic, DeepL, etc.), prompt engineering, orchestration frameworks (e.g. LangChain, LlamaIndex, or similar) - Sound judgement on when to use LLMs and when not to, and the ability to defend that call - Ability to make practical tradeoffs: cost vs quality, latency vs reliability, open source models vs other models, LLM vs deterministic systems ### Strong advantages - Hands-on experience with multi-step or AI agent-based workflows - Experience with LLM evaluation, benchmarking, and model routing layer (e.g. OpenRouter) - Experience running or self-hosting open-source models ## Who You Are - You build and ship real things for the real world. - Production systems, not demos. You work in small batches, get things in front of users, and iterate. - You have AI judgement. - You have sharp instincts for where LLMs help and where they fail, and you act on them. You can say "this shouldn't be an LLM call" or "this agent design won't hold in production" and explain why. - You think in systems, not features. - You can design how prompts, retrieval, tools, APIs, and fallbacks work together into a reliable whole across the system. - You care about the product. - You think about the person on the other end. You notice when something feels off, even if it technically works. You have strong opinions about what good AI UX looks like, and you act on them. - You have experience working in startups. - You have worked in startups before. You know the ambiguity of working with changing priorities, fast paced and with ambitions. You optimize for impact, while enjoying collaborating with a team. - You are based in the EU. - Sofia, Bulgaria is preferred, as our engineering office is there. For the right candidate, a remote EU role can be considered, but expect regular visits to the Sofia office and HQ office in Copenhagen. ## Not a Fit If - Your experience is mostly academic or research-focused - You've only built demos or prototypes - You rely heavily on frameworks without understanding their limitations ## Send your CV and one example of a production system you've shipped to [email protected]. Tell us what worked and what didn't—if it's public, include the URL. ## Questions about the role? Our CTO Georgi is happy to chat: [email protected] ## Our Recruitment Process We take hiring seriously and we know you do too. Here's what to expect: - Initial screening call: A conversation about your background, what you're looking for, and whether the role aligns with your interests. (~30 mins) - Case study: We'll send you a practical case that reflects real work you'd do here. It's not a trick - we want to see how you think about system design, tradeoffs, and production constraints. - Case presentation: You'll walk us through your approach and solution. We'll discuss your reasoning and dive deeper on decisions you made. - In-person interview: Finalists will meet the team in Sofia or Copenhagen. ## About the company Weply is a leading startup in conversational-based qualified lead generation. We currently help more than 2000 companies in Europe grow faster, by converting relevant web traffic to qualified leads through a chat solution. We have human experts man the chats on behalf of the customers and use AI under-the-hood to supercharge the experts. We also help companies provide excellent support in non-leadable conversations.