Daniel Martinez
Applied AI Engineer — Backend
Applied AI Engineer and Ph.D. candidate with experience building AI-enabled cloud services and LLM-powered workflows. Strong Python engineer with hands-on work implementing RAG and prompt engineering patterns using LangChain and Hugging Face, and integrating these capabilities into FastAPI services with authentication, logging, and evaluation-driven iteration.
Known for structured problem-solving, rapid learning, and delivering end-to-end systems that are testable, maintainable, and ready to evolve from prototype to product.
Experience
AI Solutions Developer (Consulting / Freelance)
Global (Remote)
May 2021 – Present
- LLM Workflow Integration: Built modular LLM-powered features using RAG and LangChain to automate domain-specific information retrieval and structured responses.
- Cloud-Native Backend Delivery: Developed FastAPI services packaged with Docker and deployed on AWS, designing API interfaces and integrating authentication and structured logging for reliability.
- Quality & Iteration: Ran prompt/context iterations with evaluation checks to improve response consistency and reduce failure cases.
AI Research Scientist (Visiting Researcher)
Technische Universität Dresden — Dresden, Germany
Jul 2025 – Present
- Predictive AI Modeling: Developed Deep Reinforcement Learning (DRL) algorithms for behavior prediction in complex urban environments using PyTorch and high-frequency simulations.
- Closed-Loop Validation: Designed automated testing pipelines to evaluate model robustness and decision safety in dynamic, multi-agent interactive scenarios.
AI Research Engineer (PhD Candidate)
Centro de Investigaciones en Óptica, A.C. — Aguascalientes, México
Jan 2022 – Dec 2025
- Scalable Data Infrastructure: Architected CarlaBEV, a custom simulation framework to bridge the gap between raw sensor data and semantic Bird’s-Eye View (BEV) mapping.
- End-to-End ML Pipelines: Led the design and implementation of full-stack perception and control pipelines, integrating computer vision inputs with decision-making modules.
- Computer Vision Innovation: Invented a novel data augmentation technique for semantic segmentation that significantly improved model accuracy.
Computer Vision & AI Selected Academic Publications
Front-to-Bird’s-Eye-View Transformation for Autonomous Vehicles
Published at Mexican Conference on Pattern Recognition (MCPR), pp. 166–176, 2023.
Enhancing site-specific weed detection using deep learning transformer architectures
Published in Crop Protection, 190, 107075, 2025.
Development of self-calibrating sensor footwear and relevance of in-shoe characterization
Published in IEEE Sensors Journal, 21(6), 8421–8431, 2021.
Technical Skills
Frameworks, tools, and paradigms I've mastered over the years.
Engineering & Backend
Python, FastAPI, REST APIs, Auth, Logging, Docker, Linux, Git.
LLM Engineering
Prompt Engineering, RAG, Context Design, Agentic Systems, OpenAI API, Anthropic API.
Frameworks
LangChain, Hugging Face Transformers, PyTorch, Ollama, Scikit-learn.
Data & Analytics
SQL, NumPy, pandas, Experiment tracking, Evaluation & Monitoring, Dashboards.
Cloud
AWS (S3, EC2, Lambda), Azure.
Languages
Spanish (Native), English (C1), German (B1), French (B1).