CARLOS LOPEZ MEJIA

Cloud • Data • DevOps • MLOps

Cloud & Data Engineer building adaptable platforms.

I’m Carlos, a Cloud & Data Engineer with hands-on experience delivering infrastructure-as-code, multi-environment data platforms, and batch pipelines at enterprise scale, plus a background in DevOps, IAM and security hardening, and full-stack product delivery across AWS and GCP.

View full CV Download CV (PDF)

About

Carlos has more than nine years of experience in software and data development, combining consultancy, startups, and large companies across mobility, logistics, media, and healthcare. He has led teams of 2 to 3 people and is currently expanding his technical experience and knowledge with a master's degree in Informatics at TUM Munich, focused on cloud computing, machine learning, including deep learning and natural language processing, and software engineering.

He is equally comfortable in the fast-paced environment of a startup, where he has designed data architectures from scratch, and in the more structured environment of a large company, where he has built production infrastructure at scale. That combination allows him to bring strong engineering practices to small teams without slowing down the speed they need.

He is excited by complex technical and business challenges related to collecting, interpreting, analyzing, and modeling data, and increasingly with orchestrating AI agents in real engineering workflows. He also enjoys sharing his knowledge, as he has taught classes at both personal and university levels on topics such as machine learning, MLOps, data engineering, and data analysis.

Experience

BMW: Cloud DevOps Engineer (Working Student)

Jun 2025 – Sep 2026

  • Maintain AWS infrastructure as code across production services using Terraform and AWS CDK, with staged environments and controlled rollout steps.
  • Build and support batch data pipelines with EMR/Spark, AWS Glue, Python, and Scala for enterprise-scale mobility data workloads.
  • Configure IAM roles and policies, contribute to security hardening across KMS, Secrets Manager, and VPC, and support incident and root-cause analysis.

Ocumeda: Data Engineer and Analytics (Working Student)

Jun 2024 – Jun 2025

  • Designed and implemented scalable AWS-based data architecture using S3, Lambda, Redshift, and Step Functions.
  • Delivered around 10 ETL pipelines using Python and built governance and data-quality practices for a fast-growing startup.

Selected Projects

Salsuki: Community Management Platform

2024 – Current

  • Built and maintained a full-stack platform for a non-profit salsa school with bookings, wallet balances, waitlists, and resource management.
  • Delivered the frontend with React, TypeScript, Vite, and Tailwind CSS, and the backend with Supabase, Edge Functions, Auth, and Storage.
  • Added a credit-based internal economy, event booking workflows, and interactive learning-resource features. Website: salsuki-website.vercel.app

LECture-bot: GenAI course assistant

Apr 2025 – Aug 2025

  • Designed the RAG and generative AI subsystem for a course-material assistant, processing documents, storing vectors, and integrating with cloud and local LLMs. Repository: team-LECture-bot

Skills

Cloud & Data

AWSTerraformAWS CDKEMRSparkAthenaGlueStep FunctionsQuickSight

Development & MLOps

PythonSQLFastAPIPytestDockerGitHub ActionsTerraformReactTypeScript

Education

Technical University of Munich

Oct 2023 – Expected Apr 2027

M.Sc. in Informatics with coursework in cloud computing, cloud-based data processing, DevOps, machine learning, and requirements engineering. The program has strengthened my focus on distributed systems, ML engineering, and reliable software delivery.

Instituto Politécnico Nacional

Jan 2015 – Jul 2020

B.Sc. in Computer System Engineering.

Master's Thesis

Knowledge Graph-Augmented RAG for Automotive Safety Case Generation

Expected start Aug 2026

TUM Chair of Robotics, AI and Real-Time Systems, advised by Prof. Alois Knoll and André Schamschurko. Registration is pending.

The thesis will design and evaluate a structured Knowledge Graph retrieval layer for CeCaS, replacing flat vector-based RAG to generate Ecore/OCL safety case code from UN Regulation 152, and benchmark the approach against RegulaRAG and HippoRAG using precision, recall, F1, and success or failure rate metrics.