Cloud & Infrastructure
- AWS
- EKS
- ECR
- EC2
- S3
- EFS
- Route 53
- CloudWatch
- Azure
- Linux
Senior Software Engineer at GlobalLogic
Hi, I'm Carlos Bustamante
AI & MLOps Engineer
I build production AI systems and the platforms they run on: AI agents, LLM inference with vLLM, Kubernetes workloads on EKS, and AWS environments managed with Terraform.
- AI agents
- vLLM
- AWS
- Kubernetes
- Terraform
- Ring, an Amazon company. Current client: building AI agents for internal workflows
- AWS Certified Generative AI Developer – Professional, plus 5 more certifications
01 · About
About me
AI & MLOps Engineer
I'm an AI & MLOps Engineer based in Estado de México. I build production AI systems and the infrastructure they run on: AWS environments managed with Terraform, Kubernetes clusters on EKS, containerized model services, and LLM inference with vLLM.
I'm currently a Senior Software Engineer at GlobalLogic, working with Ring (an Amazon company) on AI agents that speed up internal product and engineering workflows. Before that, at Corteza.ai, I went from fine-tuning internal LLMs to architecting the AWS and Kubernetes platform behind the company's ML and LLM workloads. I hold a B.Sc. in Mechatronics Engineering and an M.Sc. in Engineering Sciences from Tecnológico de Monterrey.
- Full name
- Carlos Alberto Bustamante Gaytán
- Current role
- Senior Software Engineer, GlobalLogic
- Location
- Estado de México, México
- Experience
- 5 years
- me@carlosbustamante.dev
- Phone
- +52 55 3131 1250
Technical skills
DevOps & IaC
- Terraform
- Kubernetes
- Docker
- AWS CDK
- Git
- CI/CD
AI/ML & MLOps
- vLLM
- LLMs
- PyTorch
- TensorFlow
- Scikit-learn
- Fine-tuning
- Model training
- Model serving
- Model deployment
Programming & Data
- Python
- TypeScript
- FastAPI
- REST APIs
- SQL
- Pandas
- NumPy
- PySpark
- Matplotlib
- Seaborn
- Plotly
Languages
- Spanish Native
- English Fluent
- German Intermediate
Certifications
Each card links to the official verification page.
02 · Background
Experience & education
From annotating AI training data to building the infrastructure that runs AI in production.
Experience
-
Senior Software Engineer Current
GlobalLogic · Client: Ring (Amazon) · Remote
– Present
- Building AI agents that accelerate internal product and engineering workflows at Ring for PMs, TPMs, and engineers.
- AWS
- Agentic AI
-
Senior Machine Learning Engineer
Corteza.ai · Remote
–
- Architected production AI infrastructure on AWS using Terraform and Kubernetes, supporting ML and LLM workloads across EKS, ECR, EC2, S3, EFS, and Route 53.
- Built Terraform-managed AMI infrastructure that cut Kubernetes pod startup time from about 5 minutes to 2 by minimizing image pull overhead.
- Migrated a production assistant from RabbitMQ-based processing to streaming inference with vLLM, raising concurrent responses from 2 to 10 on a single GPU instance.
- Terraform
- Kubernetes
- EKS
- vLLM
-
Machine Learning Engineer
Corteza.ai · Remote
–
- Led end-to-end fine-tuning workflows for internal LLMs, including codebase scraping, dataset creation, training preparation, supervised fine-tuning, and evaluation.
- Developed Python inference workflows for LLMs, object detection models, diffusion models, and third-party generative AI APIs.
- Built and deployed containerized AI services integrated with production applications and internal ML tooling.
- LLMs
- Fine-tuning
- Python
- Containers
-
Adjunct Professor
Tecnológico de Monterrey · Estado de México
–
- Taught application development, programming fundamentals, and electronics, mentoring students through hands-on micro:bit projects.
-
Data Analyst
TELUS International · Remote
–
- Analyzed and annotated digital content to improve training data quality for AI, search, and recommendation systems.
Education
-
M.Sc. in Engineering Sciences
Tecnológico de Monterrey
–
- Built a neural-network-based interface that gives students real-time feedback during lab practices.
- Co-authored research published by IEEE on an educational tool that applies data-driven control to nonlinear systems, using TensorFlow and Arduino.
-
B.Sc. in Mechatronics Engineering
Tecnológico de Monterrey
–
- Built a Flask web app that uses machine learning to estimate the time to a volcanic eruption from seismic data.
- Implemented fuzzy control for a Furuta pendulum, using MATLAB and 3D CAD.
Publication
Designing a Neural Network-Based Controller: A Practical Educational Tool for Data-Driven Control Courses
IEEE · WEEF-GEDC 2023
- Co-author. A NARMA-L2 neural network controller, trained on real pendulum data, used as a teaching tool for data-driven control courses.
03 · Portfolio
Selected projects
Cloud, machine learning, and research work. Open any project for the details.
Personal project
This site: static and bilingual on S3 and CloudFront, with a serverless contact form. All infrastructure in Terraform.
- Terraform
- CloudFront
- Lambda
OngoingView on GitHub: carlosbustamante.dev (opens in a new tab)
Kaggle competition
XGBoost classifier, tuned with Optuna, that predicts whether a loan application will be approved.
- Python
- XGBoost
- Optuna
Oct 2024View on GitHub: Loan Approval Prediction (opens in a new tab)
M.Sc. research
Desktop assistant that uses screen capture and image recognition to guide students through a motor-control lab.
- Python
- OpenCV
- TensorFlow
Jun 2024View on GitHub: PARIS: AI Learning Assistant (opens in a new tab)
Academic project
Machine learning study of the factors behind water stress in Mexico, from rainfall to aquifer levels.
- Python
- XGBoost
- LASSO
Dec 2023View on GitHub: Water Stress Analysis in Mexico (opens in a new tab)
Certificate capstone
Capstone work for the 10-course IBM Data Science Professional Certificate.
- Python
- Data analysis
- ML
Sep 2022View on GitHub: IBM Data Science Capstone (opens in a new tab)
Client project
Attendance system for an elementary school of about 300 students, built on encrypted QR ID cards.
- Python
- OpenCV
- QR codes
Sep 2022
Data science
XGBoost model for a highly imbalanced dataset: 0.96 AUC, with 87% precision and 84% recall on fraud.
- XGBoost
- Optuna
- Scikit-learn
Aug 2022View on GitHub: Credit Card Fraud Detection (opens in a new tab)
Team project
Random Forest model on INGV seismic data, served through a Flask web app that estimates the time to eruption.
- Random Forest
- Flask
- Python
Dec 2021View on GitHub: Volcanic Eruption Prediction (opens in a new tab)
04 · Contact
Let's work together
Have a role, a project, or a question about cloud or MLOps? Send me a message and I'll get back to you.