I'm Ramon Alejandro Aguirre Romero, an AI/ML Engineer and Team Lead specializing in large language models, multimodal AI systems, and computer vision. Currently leading AI development at Smarttie, where I architect enterprise conversational assistants and state-machine-driven AI solutions. I have a passion for building scalable AI products that create measurable business impact and improve accessibility.

• 🧠 Currently leading a 3-person AI team at Smarttie developing multimodal conversational assistants
• 🛠️ Former Azure Cognitive Services Support Engineer at Microsoft (via Sonata)
• 🎓 Bachelor's in Software Engineering – University of Colima, GPA: 9.7/10
• 🌟 Volunteer AI Engineer at Bootcamp Xperience creating educational AI tools

LLMs & Multimodal AI
OpenAI APIs (Assistants, Response, Whisper, TTS), Google Gemini (Fine-tuning), Anthropic Claude, Azure OpenAI, Vertex AI, RAG, Vector Databases, Function Calling, Prompt Engineering

AI/ML & Computer Vision
TensorFlow, PyTorch, YOLOv8, OpenCV, Scikit-learn, Hugging Face, LangChain, RoboFlow

Data & Analytics
Python, Pandas, NumPy, BigQuery, MySQL, SQLite, Jupyter, Sentiment Analysis

Backend & Cloud
Django, FastAPI, JavaScript, SQL, REST APIs, Azure, Google Cloud Platform, Git

Methodologies & Leadership
Team Leadership, SCRUM, CRISP-DM, Cross-functional Collaboration, Project Management

• 🧠 Lía - Emotional Support AI – Architected a sophisticated multimodal conversational agent with state-machine architecture, orchestrating multiple OpenAI APIs (Response, Assistants, GPT-4o, Whisper, TTS) for WhatsApp-based mental health support.

• 🎯 Lumi - Sales Enablement Assistant – Led development of enterprise assistant integrating GPT-4o, BigQuery, Vertex AI, and fine-tuned Gemini models. Achieved 40% reduction in response times and +12 points improvement in customer satisfaction.

• 📊 Talentto - Workplace Sentiment Platform – Built AI system analyzing employee communications with anonymization pipelines, generating HR insights on team health and conflicts.

• 🚗 Accessibility-Focused Computer Vision – Developed YOLOv8-based system detecting wheelchair and white cane users in traffic environments, achieving 90.9% mAP@50 accuracy.
  🔗 Roboflow Project

• 🛠️ Azure AI Solutions – Supported enterprise implementations of Azure OpenAI, Document Intelligence, and Language Studio, helping clients achieve 20% latency reduction and 15% cost savings.

🎓 You can also check out my University GitHub profile