FreeCourse Logo
FreeCourse.io
Verified CouponsFree CoursesJobsBlog
Categories
Home/Courses/RAG avanzado: técnicas y buenas prácticas empresariales
RAG avanzado: técnicas y buenas prácticas empresariales
IT & Software100% OFF

RAG avanzado: técnicas y buenas prácticas empresariales

Udemy Instructor
0(0 students)
Self-paced
All Levels

About this course

Este curso utiliza inteligencia artificial. Un prototipo de RAG se construye rápido. Un sistema de RAG en el que puedan confiar una clínica, un banco o una administración pública es otra cosa: tiene que encontrar la evidencia correcta, responder fiel a las fuentes, demostrar su calidad con métricas, mantenerse asequible y proteger los datos sensibles.

De eso trata este curso. Acompañamos al equipo de plataforma de Helix Medical, un sistema de salud ficticio de 4. 000 personas.

La primera versión de HelixRAG dijo que dos fármacos eran seguros juntos, aunque existía una contraindicación documentada. La recuperación había traído una guía clínica obsoleta y el modelo la repitió sin ninguna reserva: fiel a la fuente, pero errónea. De ahí nace un mandato claro: cada respuesta debe ser fiel, citada, medida y gobernada antes de la puesta en producción.

Sección a sección reconstruimos HelixRAG y medimos cada paso frente a un golden set de 300 pares pregunta-respuesta: la faithfulness pasa de 0,71 a 0,96, el context recall de 0,62 a 0,91, el costo por consulta de 0,11 $ a 0,038 $ y la latencia p95 de 6,4 s a 2,7 s. Aprenderás chunking avanzado (recursivo, por estructura, semántico, Contextual Retrieval, small-to-big), búsqueda híbrida con BM25 y Reciprocal Rank Fusion, reranking con cross-encoder, reescritura de consultas, multiconsulta y HyDE, Self-RAG y Corrective RAG, Graph RAG, recuperación multimodal sobre tablas y figuras, y los temas de producción: evaluación con un LLM como juez, eval gates en CI, optimización de costos, observabilidad y gobernanza de datos de salud según HIPAA (ley federal de EE. UU.

) y el RGPD (ley de la UE). El curso incluye siete laboratorios prácticos que se ejecutan por completo en local con Ollama, sin clave de API y sin costos recurrentes, seis tareas con soluciones de referencia, un cuestionario por sección y un examen de práctica final. Los notebooks y el código de los laboratorios están en inglés; todas las explicaciones están en español.

Skills you'll gain

Other IT & SoftwareSpanish

Available Coupons

Loading...

Course Information

Level: All Levels

Suitable for learners at this level

Duration: Self-paced

Total course content

Instructor: Udemy Instructor

Expert course creator

This course includes:

  • 📹Video lectures
  • 📄Downloadable resources
  • 📱Mobile & desktop access
  • 🎓Certificate of completion
  • ♾️Lifetime access
$0$97.99

Save $97.99 today!

Enroll Now - Free

Redirects to Udemy • Limited free enrollments

Share this course

https://freecourse.io/courses/rag-avanzado-tecnicas-y-buenas-practicas-empresariales

You May Also Like

Explore more courses similar to this one

Amazon CloudWatch: Hands-On Labs on the AWS Free Tier
IT & Software
0% OFF

Amazon CloudWatch: Hands-On Labs on the AWS Free Tier

Udemy Instructor

This course contains the use of artificial intelligence.On 15 July 2025, AWS replaced the Free Tier. It is no longer twelve months of free usage — new accounts now receive credits over a six-month window, and when that clock runs out AWS closes the account. Almost every CloudWatch tutorial online was written before that change, which means the advice they give you now quietly burns the credits keeping your account alive.This course is built around what is actually still free. CloudWatch's core allowances — 10 custom metrics, 10 alarms, 3 dashboards, 5 GB of logs — sit in the Always Free tier. They renew every month, forever, on both the free and paid plans. So every lab here runs on Lambda and SNS, which are also Always Free, instead of on EC2, which is not.You will finish with a working monitoring stack and a bill that reads zero — and we look at the bill together on screen. How the course worksEverything is built around one running company: Northwind Parcel, whose tracking API threw errors for four hours one Monday before anyone noticed. CloudWatch had been recording those errors the entire time. That gap — between data existing and somebody being told — is the whole subject of this course.Metrics — namespaces, dimensions, and the cardinality mistake that multiplied one company's monthly bill fifty-fold from a single line of codeAlarms — evaluation periods, the treat missing data as setting that decides whether a dead server pages you, and metric math that survives traffic growthLogs — retention (the default keeps charging you forever), metric filters, Live Tail, and Embedded Metric FormatLogs Insights — the highest-leverage skill in the product, taught by reading queries rather than memorising syntaxDashboards — and the discipline of deleting the widget that never changed a decisionThe agent, automation and cost — memory metrics, EventBridge auto-remediation, and how to read a CloudWatch bill How the labs workThe 12 labs are step-by-step written walkthroughs you run on your own AWS account, with every console click and every command spelled out and ready to copy. That is deliberate: you can paste a filter pattern or an IAM policy straight from the page instead of pausing a video and retyping it. Each lab states its free-tier cost up front, tells you what success looks like, and — where it matters — tells you exactly what to delete afterwards.Every lecture states what it costs before you build it. The one section that spends credits opens with a warning and ends by terminating everything on camera.

0.0•1•Self-paced
FREE$91.99
Enroll
AI-200: Certificación Azure para soluciones de IA en la nube
IT & Software
0% OFF

AI-200: Certificación Azure para soluciones de IA en la nube

Werner Ovalle

Prepárate para el examen AI-200: contenedores, datos vectoriales, integración, seguridad y un examen de práctica

0.0•10•Self-paced
FREE$84.99
Enroll
Big Data Engineering Mastery: Spark, Hadoop & Data Lakes
IT & Software
0% OFF

Big Data Engineering Mastery: Spark, Hadoop & Data Lakes

Udemy Instructor

Processing a gigabyte of data on a laptop is easy; processing a petabyte of streaming IoT data across a 100-node cluster is an entirely different profession. Welcome to the Big Data Engineering Mastery assessments! This course provides you with 200 expert-level practice questions designed to simulate the rigorous architectural and performance challenges faced by Data Engineers at top-tier tech companies.Across these four comprehensive practice exams, you will be tested on the mechanics of distributed computing. You will evaluate how to optimize Apache Spark jobs, when to implement bucketing versus partitioning, and how to resolve fatal memory errors caused by severe data skew. The questions plunge you into realistic engineering scenarios—from managing continuous supply chain inventory streams to designing financial transaction data lakes with Delta Lake.Every single question in this course is unique and includes a detailed explanation of the "why" behind the distributed architecture. By reviewing these explanations, you will learn the industry-standard methods for avoiding shuffles, applying Z-Ordering, and scaling metadata. If you are preparing for a senior Data Engineering interview or looking to validate your Big Data expertise, this is the ultimate testing ground. Enroll today and master the pipeline!Course locale: English (US)Course instructional level: Expert LevelCourse category: IT & SoftwareCourse subcategory: Data Engineering

0.0•952•Self-paced
FREE$85.99
Enroll
FreeCourse LogoFreeCourse

Freecourse.io brings you high-quality online courses with free certificates to help you upskill, boost your career, and achieve your goals anytime, anywhere.

Resources

  • Courses
  • Jobs
  • Categories
  • Features

Company

  • About
  • Blog
  • Contact

Legal

  • Privacy
  • Terms
  • Cookies
  • Licenses

© 2026 FreeCourse. All rights reserved.