Job Description
BUSCAMOS: Senior Python Cloud Engineer (Azure)
Knowledge Requirements for New Developers
1. Core Concepts
* RAG (Retrieval-Augmented Generation) — the entire system is a RAG chatbot pipeline; understanding how retrieval, chunking, embedding, and generation work together is essential
* Microservice architecture — the system is composed of ~8 independent services communicating via HTTP and message queues
2. Python
* Python 3.11 (hard requirement for Azure Functions consumption plan)
* Flask — backend, chunker, converter, connector, and image processor are all Flask apps
* FastAPI — the evaluator service uses FastAPI with async support
* Streamlit — the admin UI is built with Streamlit
* Poetry — dependency management across all services
* uv — used as an alternative Python package manager (Azure Functions)
3. Azure Cloud Services
* Azure Blob Storage — document storage, evaluation results, crawled data
* Azure AI Search (Cognitive Search) — vector/hybrid search index for document chunks
* Azure OpenAI — LLM inference (GPT-4, embeddings via text-embedding-3-large)
* Azure Document Intelligence (Form Recognizer) — PDF/document parsing and OCR
* Azure Container Apps — production hosting for all microservices
* Azure Container Registry (ACR) — Docker image storage and build
* Azure Functions — queue-triggered indexing pipeline (blob → convert → chunk → embed → index)
* Azure Queue Storage — async messaging between services (indexing queue, poison queue)
* Azure EntraID — authentication and role-based access control
* Azure Application Insights — telemetry and logging
* Azure CLI — deployment, ACR interactions, Container App updates
4. Docker & Containers
* Docker — every service has a Dockerfile; local development and production both use containers
* Building, running, and debugging containerized Python services
5. Data Sources & Crawling
* SharePoint Online — document library crawling with auth
* Confluence — page and page-tree scraping with PATs
* Web crawling — generic website crawling with depth control
6. AI / ML Evaluation
* RAGAs framework — evaluation metrics (faithfulness, answer relevance, context precision/recall)
* LLM-as-judge — using GPT-4 to score answer quality
* Locust — load/performance testing framework for the RAG API
7. Observability
* LangFuse — LLM observability and tracing platform
* Application Insights — Azure-native telemetry
8. Development Tools
* VS Code — launch configurations, debug configs, tasks
* Azurite — local Azure Storage emulator (queues, blobs)
* Azure Functions Core Tools — local Azure Functions runtime
* Git — branching model: dev/ → dev/featureX → PR to main
Idioma
■ Español
■ Inglés C1
·
OFRECEMOS:
· Contrato en modalidad Freelance Full time
· Oferta económica: 250€/290€ jornada + IVA (según experiencia aportada)
· Proyecto de larga duración
· Localización: España (Teletrabajo 100% )