🤖 LLMs & AI Agents
Agents and assistants that reason, call tools and act on real business systems, built with MCP and function calling.
Hey, I'm Martí Grau
> AI Lead @ REVEL
AI Lead at REVEL, turning LLMs and AI agents into real products that move the business. Telecom engineer and co-founder of StageInHome, working with generative AI since 2019, from diffusion models and GANs to production LLM systems on the cloud.
- 7+
- years building AI
- 1
- AI startup co-founded
- 3
- AI leadership roles
~/marti/agent.py
01 // what I do
AI capabilities
📚 RAG & Knowledge
Retrieval pipelines that ground LLMs on company data, with evaluation loops to keep answers accurate.
🎨 Generative Vision
Diffusion models, GANs and conditioning strategies for controllable, spatially aware image generation.
☁️ MLOps & Cloud
Dockerized services, auto-scaling inference and parallel training on AWS (SageMaker, Lambda, ECS, S3).
🧪 Fine-tuning & R&D
Model scouting, fast experimentation and fine-tuning on domain data to get the most out of state-of-the-art models.
🧭 AI Strategy & Leadership
AI roadmaps tied to business value, multidisciplinary teams, and turning complex systems into clear decisions.
# Python # PyTorch # Hugging Face # Diffusers # LLM APIs # AI Agents # MCP # RAG # Vector DBs # Fine-tuning # Prompt Engineering # Evals # Python # PyTorch # Hugging Face # Diffusers # LLM APIs # AI Agents # MCP # RAG # Vector DBs # Fine-tuning # Prompt Engineering # Evals
# AWS # SageMaker # Lambda # ECS # ECR # S3 # Docker # Google Cloud # FastAPI # Data Pipelines # Auto-scaling # CI/CD # AWS # SageMaker # Lambda # ECS # ECR # S3 # Docker # Google Cloud # FastAPI # Data Pipelines # Auto-scaling # CI/CD
02 // where I've been
Experience
- View details
- 🧭 Own the AI product vision, roadmap and long-term strategy at REVEL, one of Spain's leading car renting and subscription companies.
- 🤖 Lead the design and delivery of LLM-powered agents and assistants that automate customer experience and internal operations.
- 🏗️ Design and build end-to-end AI infrastructure, from data pipelines to scalable cloud architectures with auto-scaling and production reliability.
- 🔌 Connect AI with the company's systems and data through tool integrations and MCP servers, so every team can use AI safely in its daily work.
- 🌱 Drive AI adoption and culture across the organization, prioritizing use cases by business impact.
- View details
- 🎯 Direct the strategy and technical execution for cutting-edge Generative AI solutions across diverse industrial sectors.
- 👥 Coordinate a multidisciplinary team to define, architect, and deploy scalable AI system solutions, spanning data pipelines, model orchestration, and production inference.
- ⚙️ Manage the full lifecycle of AI products, from hands-on R&D to scalable production deployment.
- 🚀 Drive the adoption of state-of-the-art GenAI models and AI tooling through continuous model scouting, rapid experimentation, and fine-tuning on domain-specific data to maximize real-world impact.
- View details
- 🏗️ Founded and led the end-to-end AI pipeline, from dataset design and curation to model training, validation, and production deployment.
- 👁️ Designed and trained Generative Computer Vision systems, focusing on diffusion models and conditioning strategies for controllable, spatial-aware image generation.
- 🔌 Built and maintained production-ready APIs enabling seamless integration between AI models and web-based products.
- ☁️ Architected and operated AI infrastructure on AWS, including parallel training, Dockerized services, and auto-scaling inference pipelines.
- 📚 Acted as the technical GenAI reference, tracking state-of-the-art models and driving informed adoption across the team.
- 🎤 Represented the technical vision in startup accelerators and investor pitches, translating complex systems into clear business value.
- 🤝 Mentored and onboarded junior AI engineers, defining development standards and fostering strong engineering practices.
- View details
- 🔬 Researched and developed novel Generative Adversarial Networks (GANs) for image style transfer within indoor environments.
- 🧪 Successfully prototyped the core technology that led to the founding of the startup StageInHome.
- 💡 Developed and integrated multiple machine learning models for internal innovation projects.
- View details
- 🍳 Contributed to the Smart Recipe Hub project for Nestlé, handling technical feature development and platform troubleshooting.
- 🤖 Developed and deployed a custom chatbot leveraging Python and Google Dialogflow services.
🎓
2019 - 2021
Master's in Advanced Telecommunication Technologies
Deep Learning for Multimedia Processing · UPC, ETSETB TelecomBCN
03 // who I am
About me
I'm Martí, a Telecommunications Engineer from Barcelona who has been working with generative AI since 2019, long before it was everywhere. I started doing research on GANs for interior design, and that prototype turned into StageInHome, the startup I co-founded and where I led all the AI for almost five years.
Along the way I've trained diffusion models, built production APIs and set up auto-scaling AI services on AWS. Later, as AI Lead at Napptilus Tech Labs, I led a multidisciplinary team shipping LLM and RAG solutions for clients in many different industries.
Today I'm AI Lead at REVEL, where I define the AI strategy and build agents and AI infrastructure that people actually use every day. What I enjoy most is finding the simplest AI solution that has a real impact, and then making it reliable in production.
When I'm not building, I write about AI on my blog.
04 // let's talk
Have an AI idea in mind?
I'm always happy to talk about generative AI, agents and how to take them to production. Send me a message on LinkedIn!