Acerca de Rafael
- Agentic Workflows & Autonomous Agents: Implementing advanced reasoning loops using LangGraph, CrewAI, and AutoGen. I build systems that don't just talk—they execute tasks, use tools, and make informed decisions.
- Advanced RAG Systems (Retrieval-Augmented Generation): Architecting robust data pipelines with Vector Databases (Pinecone, Milvus, or FAISS). I solve the "hallucination" problem to ensure your LLM provides precise, data-driven answers based on your private documentation.
- Production-Ready AI Backends: High-performance APIs built with Python (FastAPI) and Node.js, engineered to scale and optimized for token consumption and low latency from day one.
- AI-Powered E-commerce & Marketplaces: Enhancing complex platforms (Next.js Commerce, SaaS) with semantic search, automated product tagging, and intelligent recommendation engines.
- AI/ML: SOTA Models (OpenAI, Anthropic Claude, Google Gemini), Open-source LLMs (Llama, Mistral), LangChain, LangGraph, CrewAI.
- Vector Ops: Pinecone, Milvus, Weaviate, FAISS.
- Backend & Data: Python (FastAPI), Node.js, Docker, SQL/NoSQL.
- Hands-on: I am an active developer, not just a consultant.
- Cost-Conscious: I architect for ROI, optimizing LLM calls to prevent runaway API costs
- Reliable: Clean code, >90% test coverage, and a focus on long-term maintainability
Inglés
Competencia profesional completa
Español
Bilingüe o nativo
Catalán
Competencia profesional básica
Experiencia
- Stealth Mode StartupSenior Generative AI Developer (Contractor / Stealth Project)julio de 2025 - Hoy (11 meses)Currently architecting and deploying production-grade Generative AI solutions for high-scale environments. My focus is on moving beyond simple prompting into complex Agentic Workflows and specialized multi-layered AI architectures.Key Technical Achievements:
- Multi-Agent Orchestration: Designed and implemented an Orchestrator-Worker model using LangGraph and LangChain. This system manages specialized sub-agents to automate complex internal decision-making processes and long-running business workflows.
- Autonomous Research Agents: Developed agents integrated with Tavily for real-time internet search, capable of autonomous data synthesis and automated generation of dynamic business outputs, including PPTX presentations, Excel reports, and Miro boards.
- Advanced RAG Pipelines: Engineered robust Retrieval-Augmented Generation (RAG) systems using Pinecone and FAISS. Optimized retrieval accuracy through custom embedding strategies and re-ranking to eliminate hallucinations in technical documentation queries.
- LLM Ops & Backend Scalability: Orchestrated high-performance asynchronous APIs with FastAPI to handle streaming responses. Implemented strict monitoring for token consumption and model evaluation to ensure cost-effectiveness and reliability.
- Port AventuraSenior fullstack developerENTRETENIMIENTO & TIEMPO LIBREjulio de 2023 - mayo de 2025 (1 año y 10 meses)Tarragona, EspañaPlayed a key role in architecting and developing a high-performance custom CMS, specifically designed to handle large-scale content operations. My focus was on bridging the gap between robust backend architecture and seamless editorial workflows.Key Contributions:
- High-Scale Architecture: Engineered a custom, modular CMS infrastructure capable of managing thousands of dynamic assets, ensuring low latency and high availability for global traffic.
- Performance Optimization: Spearheaded the transition to a Headless architecture, significantly improving content delivery speeds and multi-platform integration capabilities.
- Workflow Automation: Designed and implemented automated content validation pipelines and SEO optimization tools, reducing manual editorial tasks by 30%.
- Tech Stack: Leveraged Python (FastAPI), PostgreSQL, and Redis to build a resilient and scalable backend, following strict Clean Architecture and SOLID principles.
- Full stack developingTech Lead en React & developer PythonHIGH TECHenero de 2022 - junio de 2023 (1 año y 4 meses)Lead the hands-on architecture and technical direction of high-security financial platforms, managing a specialized team of 3 engineers to deliver robust, mission-critical systems.Key Technical Achievements:
- Security-First Architecture: Designed and implemented financial solutions compliant with strict security standards, ensuring end-to-end data encryption and secure transaction processing.
- Technical Leadership (Hands-on): Managed a high-performing engineering team through the full SDLC, balancing active coding with architectural decision-making and mentorship.
- Scalable Infrastructure: Engineered modular backend systems using Python/FastAPI and PostgreSQL, optimized for high-concurrency and financial data integrity.
- Code Quality & Compliance: Implemented rigorous testing frameworks and automated CI/CD pipelines, maintaining >90% code coverage to meet financial audit requirements.
Recomendaciones
Estos perfiles de freelance también coinciden con tus criterios
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Formación
- Mastering Generative AI: Agents with RAG and LangChainIBM2025Especialización en creación de agentes de IA inteligentes usando RAG (Retrieval-Augmented Generation) y LangChain. Domino la construcción de chatbots empresariales, sistemas de búsqueda semántica y asistentes virtuales personalizados
- Python for AI & Development ProjectIBM2025Proyecto práctico intensivo combinando Python con desarrollo de aplicaciones de inteligencia artificial. Experiencia hands-on en machine learning, procesamiento de datos, APIs de IA y deployment de soluciones escalables en entornos de producción