I build production-ready AI systems, scalable software architectures, and intelligent automation workflows that transform complex business problems into reliable, high-impact products.

My work combines advanced AI engineering, software architecture, and deep technical research with a strong focus on scalability, maintainability, and real-world deployment.

• Design and deploy LLM-based systems, RAG pipelines, and multi-agent architectures
• Build scalable backend systems and AI-integrated web platforms
• Develop automation ecosystems for enterprise workflows
• Lead technical teams and mentor developers in AI engineering and software architecture
• Research next-generation Transformer architectures and computational optimization methods

Alongside my engineering career, I conduct independent PhD-level research focused on the internal mechanics, scalability, and efficiency of Transformer architectures and Large Language Models.

• Transformer optimization and computational efficiency
• Internal representation dynamics in deep architectures
• Asymptotic dimensional contraction in attention systems
• Efficient training dynamics for large-scale language models
• Mathematical modeling of deep learning systems
• Stability and scalability of long-context architectures

OmegaMGT: Isometric Dynamics and Asymptotic Dimensional Contraction in Single-Highway Transformers

A novel Transformer architecture designed to drastically improve computational efficiency, scalability, and training stability in large language models through mathematically grounded structural modifications.

I specialize in building systems that move beyond prototypes into robust production environments.

• AI Architecture
• Multi-Agent Systems
• Retrieval-Augmented Generation (RAG)
• Backend Engineering
• Software Architecture
• Automation & Orchestration
• AI Reliability & Evaluation
• Scalable System Design
• Production AI Deployment

• Led AI initiatives as Technical Lead & AI Architect for enterprise environments
• Built AI systems capable of reducing workflows from days to minutes
• Developed AI-powered legal drafting systems with high reliability
• Architected marketing automation ecosystems using generative AI
• Built production-ready AI assistants with LangChain, Supabase, and RAG
• Mentored students and developers in Computer Vision and Generative AI

B.Sc. in Systems Engineering
Universidad de Antioquia

Technical Diploma in Software Development
Universidad Autónoma de Bucaramanga

“Building AI systems that are not only intelligent, but scalable, reliable, and grounded in deep technical understanding.”