ABOUTapeirum helps ambitious companies overcome delivery bottlenecks by providing high-performing engineering teams in Spain and pairing them with expert digital product design. We specialize in scalable nearshore engineering teams, UX/UI design, and practical AI solutions to support real workflows. Our focus is on fast, efficient, and modern implementation processes, emphasizing clear communication, product understanding, and ownership. With multilingual expertise, we deliver custom solutions to drive confidence, efficiency, and long-term success for our clients.Currently we're seeking support with one of our clients, an innovative start-up to build the factories of tomorrow. They're redefining "smart" for manufacturing SMEs with Industrial Io T, Data, AI and cloud—moving from visibility, to optimization, to LLM-powered agents that continuously improve production.In this position you'll build LLM-based workflows and agentic features for real factories: ship clean services in Python/C#, wire up retrieval (embeddings + vector/hybrid search), add tool/function calling so models can act, and iterate with domain experts. We value self-driven engineers who communicate transparently ("I don't know yet—can we figure it out together?" beats over-promising).Location: Spain (remote-first). Preference for North of Spain (Galicia, Asturias, Cantabria, Basque Country). No profiles outside Spain. Occasional travel to Switzerland.Process: Intro → team chat → tech deep-dive → short practical → final chat.YOUR TASKS
Aumente sus posibilidades de conseguir una entrevista leyendo la siguiente descripción general de este puesto antes de presentar su candidatura.
Build services & APIs in Python and C# (. NET), integrating with Io T/ERP/MES and cloud components (Azure preferred).Design & implement LLM workflows and agents: tool/function calling, retrieval pipelines, prompt/JSON orchestration, and safe control flow (graphs/state).Own quality end-to-end: tests, reviews, CI/CD, telemetry, tracing/logging; guardrails and simple evals; keep things observable and debuggable.Evaluate AI in production: add offline/online evaluation, benchmarks, and safety checks; measure value, not just accuracy.Collaborate transparently: give realistic estimates, surface risks early, and ask for help when needed.Continuously improve: iterate with domain experts to turn prototypes into stable, maintainable services. In your first 90 days you might:
Ship a small Python or C# service that talks to an ERP/MES.Launch an agentic workflow (2 tools, JSON I/O) with retries, tracing/metrics, lightweight evals & guardrails.Productionize (CI/CD, auth, alerts); add knowledge access only if needed (RAG or deterministic) and track value (success rate, p95). WHAT YOU'LL BRING
3–5 years of software engineering experience (or equivalent impact).Proven hands-on experience with Claude Code in production or advanced prototyping environments.Strong orchestration mindset: ability to design systems, workflows, and AI-driven processes by coordinating tools, agents, and engineers — going beyond pure feature implementation.Actively explores emerging technologies and AI advancements, evaluates their practical relevance, and drives structured adoption within the team.Strong Python and C#/. NET skills; you've built and shipped APIs/services.You've shipped at least one LLM product or serious POC (cloud or open-source models).Experience with at least one LLM orchestration framework (e.g., Lang Graph/Lang Chain—others welcome).Solid with datastores (SQL/No SQL) and event/streaming basics.Hands-on with RAG and a vector/hybrid search backend (e.g., Postgres/pgvector, Qdrant/Weaviate, Elasticsearch/Azure AI Search).Solid with SQL (schemas, migrations); working knowledge of No SQL/event/streaming basics.Cloud & containers: Docker; CI/CD (Git Hub Actions/Azure Dev Ops); basic Kubernetes/Azure or willingness to learn quickly.Mindset: self-driven, outcome-oriented, and radically transparent in communication.Language: English (min. C1) Nice to have:
Industrial/edge exposure (OPC UA, MQTT), queues/streams, or time-series data.ML engineering basics (Py Torch/Tensor Flow), feature stores, data pipelines.Observability (Open Telemetry, Grafana), security, and role-based access in data/AI systems. HOW WE WORK (AND WHY IT'S FUN)
AI‐first, pragmatism‐driven: Use AI where it makes a real difference; use classic engineering where it's better.Small team, big ownership: You'll have autonomy, clear goals, and direct impact on customers.Real industry impact: Your code helps reduce waste, save energy, and make factories more resilient. READY TO MAKE AN IMPACT?We're looking for someone who thrives in a fast-paced, innovation-driven environment — someone adaptable, curious, and ready to solve real problems with smart, modern tools. If that sounds like you, we'd love to hear your story. xpzdshu Send your CV and 1–2 links (Git Hub, demo, or brief write-up of an LLM workflow you built).