> Markdown version of [/jobs/ext/2128638-principal-ai-full-stack-engineer](https://www.wearedevelopers.com/jobs/ext/2128638-principal-ai-full-stack-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Principal Ai Full-Stack Engineer - **Company:** Peak3 (Formerly Za Tech) - **Location:** Madrid, Spain - **Contract:** Permanent contract - **Skills:** Artificial Intelligence, Software as a Service, Cloud Engineering, Software Debugging, Distributed Systems, Python (Programming Language), PostgreSQL, Next.js, Software Safety, TypeScript, AI Infrastructure, ReactJS, Large Language Models, Multi-Agent Systems, Prompt Engineering, Backend, AI Platforms, Kubernetes, Free and Open-Source Software, Machine Learning Operations, Front End Software Development, Serverless Computing - **Published:** August 19, 2026 - **Apply:** https://www.buscojobs.com.es/principal-ai-full-stack-engineer-en-alcobendas-ID-367838328 ## About the Role operating production software at scale, ideally including an LLM/AI system taken from prototype to production.Fluent in Python or TypeScript and a modern frontend framework, with genuine comfort spanning backend, frontend, infrastructure, data, and model code.Hands-on LLM engineering depth: RAG/retrieval, function/tool calling, agent frameworks (LangGraph, LlamaIndex, or custom), systematic evals, prompt- and model-level optimization.Solid AI infrastructure grounding - vector databases, embedding models, model deployment/serving, inference optimization, distributed systems - with the operational discipline for regulated environments.Strong product instincts: can turn ambiguous problems into well-scoped solutions, cares about real user impact, and argues for simplicity when complexity isnt earned.High agency - prototypes and drives progress without waiting for complete specs.Nice to HaveTaken an LLM system from 0 to 1 - prototype to real users, revenue, and operational load, including on-call and incident reviews.Agent infrastructure experience: secure execution sandboxes (gVisor, Firecracker, WASM), agent memory/state architectures, model routing, workflow orchestration.External technical signals - open-source contributions, technical writing, conference talks.Interest in multimodal AI, long-context engineering, applied AI safety/alignment, or RLHF.Finance or insurance domain experience (engineering ability and learning velocity matter more). Our stack: Python, TypeScript, React/Next.Js, Postgres, Kubernetes, and various vector databases and LLM providers. Prior experience with every part isnt required - strong fundamentals and fast learning matter more. ## Description About Peak3Peak3 is an award-winning vertical SaaS provider, enabling more relevant, convenient, and affordable insurance protection for everyone through our technology and ingenuity.Together with our clients, we create a more resilient and innovative future.We combine insurance core, distribution, and AI solutions to deliver a step change in performance for insurers, MGAs, and insurance intermediaries.From greenfield embedded insurance ventures to multi-country core modernization programs, our SaaS solutions power top customers across life, health, and P&C insurance.Our 500+ colleagues are based across over 15 countries in Europe, Asia and the Middle East - with an ambitious roadmap to scale further.About The RoleAs a Principal AI Full-Stack Engineer on the AI Platform team, youll own the design, build, and operation of the agentic AI systems powering our core insurance products - architecting LLM inference chains, building production AI infrastructure, shipping full-stack features from model to UI, and setting technical direction for the team.You should be as comfortable debugging a RAG recall failure in production as whiteboarding agent memory architectures with research or debating interaction patterns with design.ResponsibilitiesBuild agentic AI applications end-to-end - RAG pipelines, multi-agent orchestration, tool calling, workflows - and ship insurance agents that run in production.Design and optimize LLM inference chains: prompt engineering, structured outputs, fine-tuning (LoRA/PEFT), evals, and observability.Stand up production-grade AI infrastructure: vector databases, semantic retrieval, model gateways, inference acceleration, and cost governance.Own full-stack delivery - frontend (React/Next.Js), backend (Python/TypeScript), cloud-native deployment (K8s/Serverless) - taking demos to policy volumes in the millions.Partner with actuarial, underwriting, and claims experts to turn domain knowledge into trustworthy, explainable, auditable AI systems.Experience & QualificationsProven track record shipping and operating production software at scale, ideally including an LLM/AI system taken from prototype to production.Fluent in Python or TypeScript and a modern frontend framework, with genuine comfort spanning backend, frontend, infrastructure, data, and model code.Hands-on LLM engineering depth: RAG/retrieval, function/tool calling, agent frameworks (LangGraph, LlamaIndex, or custom), systematic evals, prompt- and model-level optimization.Solid AI infrastructure grounding - vector databases, embedding models, model deployment/serving, inference optimization, distributed systems - with the operational discipline for regulated environments.Strong product instincts: can turn ambiguous problems into well-scoped solutions, cares about real user impact, and argues for simplicity when complexity isnt earned.High agency - prototypes and drives progress without waiting for complete specs.Nice to HaveTaken an LLM system from 0 to 1 - prototype to real users, revenue, and operational load, including on-call and incident reviews.Agent infrastructure experience: secure execution sandboxes (gVisor, Firecracker, WASM), agent memory/state architectures, model routing, workflow orchestration.External technical signals - open-source contributions, technical writing, conference talks.Interest in multimodal AI, long-context engineering, applied AI safety/alignment, or RLHF.Finance or insurance domain experience (engineering ability and learning velocity matter more).Our stack: Python, TypeScript, React/Next.Js, Postgres, Kubernetes, and various vector databases and LLM providers.Prior experience with every part isnt required - strong fundamentals and fast learning matter more. ## Related Videos - [Developing the Backend with Stefan Lingler, CTO at Shpock](https://www.wearedevelopers.com/videos/100360-developing-the-backend-with-stefan-lingler-cto-at-shpock) - [Watch Tests Go Brrrr! : Getting Started with Cypress in ReactJS](https://www.wearedevelopers.com/videos/282-watch-tests-go-brrrr-getting-started-with-cypress-in-reactjs) - [GraphQL + Apollo + Next.js: A Lovely Trio](https://www.wearedevelopers.com/videos/311-graphql-apollo-next-js-a-lovely-trio) - [One AI API to Power Them All](https://www.wearedevelopers.com/videos/1601-one-ai-api-to-power-them-all) - [Nest.js - TypeScript in the backend can also be clean](https://www.wearedevelopers.com/videos/1033-nest-js-typescript-in-the-backend-can-also-be-clean) - [Bringing the power of AI to your application.](https://www.wearedevelopers.com/videos/1010-bringing-the-power-of-ai-to-your-application) ## Related Articles - [What Are Large Language Models?](https://www.wearedevelopers.com/magazine/304-what-are-large-language-models) - [MLOps And AI Driven Development](https://www.wearedevelopers.com/magazine/82-mlops-and-ai-driven-development) - [Graph and AI Trends 2026: Why Is AI Running but Not Yet Delivering?](https://www.wearedevelopers.com/magazine/680-graph-and-ai-trends-2026-why-is-ai-running-but-not-yet-delivering) - [How to Become an AI Engineer](https://www.wearedevelopers.com/magazine/331-how-to-become-an-ai-engineer) - [Dev Digest 137 - AI'm not sure about this](https://www.wearedevelopers.com/magazine/485-dev-digest-137-ai-m-not-sure-about-this) - [Dev Digest 121 - AI goes offline](https://www.wearedevelopers.com/magazine/456-dev-digest-121-ai-goes-offline)