Live online classes
10 ECTS
Expert Programme in LLM Engineering and Agentic AI
- Prior knowledge is required.
- We prepare you to earn certifications.
Get in touch
Why study the Expert Programme in LLM Engineering and Agentic AI?
There is plenty of content explaining what a transformer is or how to write a prompt. What is far less common is training that takes you from theory to a deployed product, with verifiable quality metrics and a real cost analysis. Here, each module ends with a functional deliverable, not a multiple-choice exam.
The final project is a complete system that combines a production-grade RAG solution with an autonomous agent, deployed in the cloud with a public URL, an observability dashboard and inference cost control. It is the kind of project that carries real weight in an AI technical interview.
This Master’s programme also prepares you to:
- Prepare for the AWS AI Practitioner (AIF-C01) and NVIDIA NCA-GENL certifications, with mock exams included in the programme.
- Build a portfolio with six projects published on GitHub and real deployed systems that you can showcase from day one of your job search.
- Progress from developer to LLM Engineer or AI Engineer, a career path with strong prospects and sustained demand.
Prerequisites
The Master’s programme is designed for developers and technical professionals who want to specialise in LLM engineering. It is suitable both for those who already code in Python and want to pivot into AI, and for data or ML professionals looking to master the production stack for language models. No prior experience with LLMs or AI frameworks is required.
Mandatory Requirements
- Verifiable knowledge and at least one year of professional experience in full-stack programming (Python, Java, JavaScript or related languages).
- Machine learning fundamentals: what a model is, training, validation, overfitting and evaluation metrics.
- REST API experience: HTTP requests, JSON format and API key authentication.
- Operational development environment: VS Code or Jupyter, and basic Git (clone, commit, push).
Recommended Requirements:
- An introductory deep learning course (fast.ai, Coursera or another equivalent course).
- Familiarity with NumPy and pandas for data manipulation.
- Knowledge of SQL and basic experience with a cloud provider (AWS, GCP or Azure).
- Experience using ChatGPT, Claude or another LLM in a technical or professional context.
What will you learn?
- Build production-grade RAG systems, from document ingestion through to quality evaluation with RAGAS.
- Design autonomous agents with LangGraph and Model Context Protocol (MCP), with memory, tool use and cost control.
- Apply advanced prompt engineering: few-shot prompting, chain-of-thought, function calling and structured outputs.
- Operate LLMs in production: observability, continuous evaluation, cost optimisation, and deployment with Docker and in the cloud.
- Apply Responsible AI: bias detection, transparency, guardrails, and the EU AI Act and OWASP LLM Top 10 frameworks.
- Prepare for the AWS AIF-C01 and NVIDIA NCA-GENL certifications.
Tools
Learn how to use industry-leading tools
Certification training
The Master’s programme covers the syllabus for two internationally recognised certifications and integrates them throughout the programme, with mock exams in the final weeks:
- AWS Certified AI Practitioner (AIF-C01): AI and machine learning fundamentals on AWS, foundation models, RAG, Responsible AI, security and governance. Covered throughout Modules 1 to 5.
- NVIDIA Certified Associate, Generative AI LLMs (NCA-GENL): prompt engineering, RAG systems, LLM evaluation and metrics, bias detection, and containerised deployment. Covered across Modules 2 to 6.
Preparation is integrated into the programme, with mock exams for both certifications. Registration for the official AWS and NVIDIA exams is managed separately; the admissions team will provide you with the relevant details. Passing the exams is not guaranteed, but the programme content and practical work are aligned with the objectives of each certification.
Curriculum for the Expert Programme in LLM Engineering and Agentic AI
The programme is structured around five modules and a final project (72 synchronous hours / 240 total hours / 10 ECTS). Each module begins with a challenge that you do not yet know how to solve on day one, while the live sessions take the form of guided, hands-on workshops.
The programme includes 72 hours of synchronous instructor-led teaching and requires a total of 240 hours of student study, including projects, practical work, labs and self-study.
Assessment: 20% notebooks and exercises per module + 20% module challenges + 10% peer review + 10% AWS AIF-C01 mock exams + 10% NVIDIA NCA-GENL mock exams + 30% Capstone.
*The academic program may be subject to changes in line with the changing demand for specific skills in the market.
Module 1 – 0,5 ECTS
LLM and API Fundamentals
Transformer architecture, tokenisation and generation parameters (temperature, top_p, penalties). An overview of current models (GPT-4o, Claude, Gemini, Llama 3, Mistral) and an introduction to Amazon Bedrock and the OpenAI and Anthropic APIs.
Module 2 – 1,5 ECTS
Advanced Prompt Engineering
From zero-shot to few-shot prompting, chain-of-thought, function calling and structured outputs with Pydantic and Instructor. Prompt evaluation using test datasets and A/B testing, prompt management and versioning, and bias and hallucination detection.
Module 3 – 2 ECTS
RAG Systems and Evaluation
Complete RAG architecture: chunking, embeddings, vector databases (Pinecone, Qdrant, pgvector) and reranking. Advanced techniques (HyDE, multi-query, self-RAG), rigorous evaluation with RAGAS, and a full-stack application built with FastAPI and Next.js.
Module 4 – 2 ECTS
AI Agents and Orchestration
Agent patterns (ReAct, plan-execute, reflection), LangGraph for state graphs and Model Context Protocol (MCP). Multi-agent systems, memory management, Amazon Bedrock Agents, and evaluation of reliability, costs and security guardrails.
Module 5 – 2 ECTS
LLMOps, Production and Responsible AI
AI-specific observability (Langfuse, LangSmith, Arize Phoenix), continuous evaluation and cost optimisation through caching and routing. Deployment with Docker, vLLM and Ollama on AWS, together with Responsible AI, security, governance and safety in production.
Capstone – 2 ECTS
Production RAG System and Agent
You will design, implement and deploy a complete product that combines a production-grade RAG system with an autonomous agent capable of orchestrating multiple tools. The system is deployed in the cloud, monitored using LLMOps, and delivered with a real-world analysis of inference costs, which you will present and defend before an assessment panel through a live demo and code review.
Career opportunities
The LLM Engineer profile has one of the largest gaps between supply and demand in today’s tech market. Upon completing the programme, you can pursue roles such as:
- LLM Engineer / AI Engineer
- RAG Systems Engineer / Knowledge Engineer
- AI Agents Developer / Agentic Systems Engineer
- AI Product Engineer
- AI Consultant or Freelancer for Projects
According to the 2026 Salary Guide published by Revista Inteligencia Artificial, LLM Engineer profiles working in production in Spain earn around €68,000–€72,000 per year. Get on Board recorded a 340% increase in demand for LLM/AI Engineer profiles in Latin America between 2023 and 2025. Remote work for North American companies is common in this field, which broadens the market for professionals in Latin America.
Career Readiness
During the programme, you will have access to IMMUNE’s Career Readiness service: a personalised employability pathway that includes support with preparing your technical CV and LinkedIn profile, mock technical interviews, connections with companies in the IMMUNE ecosystem, and access to our job board.
The aim is for you to finish the programme with solid knowledge and be able to demonstrate it during a recruitment process. The six deployed projects and the certifications you prepare for are among the elements that carry the most weight in an AI technical interview.
A comprehensive training experience
The programme includes a Human Sciences component: skills that complement your technical profile. In AI engineering, particular emphasis is placed on communicating technical decisions to business stakeholders, analysing the return on investment of a system, and making decisions under uncertainty. Being able to justify why a system costs what it does, or why an architecture scales, can make a real difference in product and technical leadership roles.
FAQs about the Expert Programme in LLM Engineering and Agentic AI
What qualification will I receive upon completion?
Upon completing the programme, you will receive the IMMUNE Technology Institute Diploma in LLM Engineering and Agentic AI, a proprietary, non-regulated qualification recognised within the tech industry.
Do I need prior experience in Artificial Intelligence?
No prior experience with LLMs or AI frameworks is required. You do need intermediate Python skills, a foundation in machine learning, and experience working with REST APIs. From that starting point, the programme takes you through to production-ready systems.
Are the AWS and NVIDIA certifications included?
The programme prepares you for both certifications, covers their syllabuses and includes mock exams. Registration for the official AWS and NVIDIA exams is managed separately; the admissions team will provide you with the relevant details. Passing the exams is not guaranteed, but the programme content is aligned with each exam.
What technical requirements does my computer need to meet?
A laptop with a camera and microphone, at least 8 GB of RAM, and an i5 processor or equivalent. A stable internet connection is recommended for live classes and lab sessions.
Is it 100% online?
Yes. Classes are delivered live with the group and are recorded so you can review them afterwards. Lab work and assignments are completed independently.
What is the Capstone Project?
It is the final project: a system that combines a production-ready RAG solution with an autonomous agent, deployed in the cloud with a public URL, an observability dashboard and cost analysis. You will present and defend it before an assessment panel with a live demo. It forms part of the portfolio you can showcase in interviews.
Is it compatible with a full-time job?
Yes. The format consists of two live three-hour sessions per week, plus independent work that you can organise at your own pace. The final project is concentrated in the final weeks.
Is there a career guidance service?
Yes. All students have access to IMMUNE’s Career Readiness service, which includes CV preparation, career guidance, mock interviews and access to our job board.
Are scholarships available?
Financing options and discounts are available for certain profiles. Our admissions team can provide information on the options available at the time of your application.
Financing
Full payment
Check the discounts available for upfront payment.
BBVA
If you are a resident in Spain, you can finance your programme through BBVA.
Sequra
Pay in installments, even if you are unemployed and cannot guarantee the loan.
Quotanda
Pay in installments, even if you are unemployed and cannot guarantee the loan.
Fundae
Pay for your training through the Spanish Employment Training Foundation. Aimed at active workers who wish to finance their program through the subsidized training program.
Book a personalised academic consultation
Flor Biscardi
Agustina Ruíz
Admissions Process
Our students are characterized by their passion for technology. Our admissions process focuses on who you are, how you think, what you have accomplished, and then sharing your goals.
Our aim is to get to know you better, see what makes you unique and ensure that the IMMUNE educational model adapts to your profile.
English (UK)