Ernesto Aguilera

Senior Cloud DevOps & AI Platform Engineer with 8+ years of experience designing secure, scalable, cloud-native solutions across AWS, Azure, and Google Cloud for enterprise AI/ML products. Currently at Teradata R&D, architecting a centralized AI platform standardizing LLM/ML access for 100+ enterprise users — managing 400+ microservices on Kubernetes (EKS, AKS, GKE, Rancher) and reducing cross-team incident resolution by 97% (4h to 30min). Proven expertise in LLMOps, agentic AI workflows (LangChain, LangGraph, MCP), multi-cloud GitOps (Flux, Argo, GitHub Actions), and DevSecOps (FIPS, NIST, mTLS, Keycloak, RBAC/ABAC).

Polyglot engineer fluent in Python, Java, Go, Rust, C/C++, and Elixir. UNAM Computer Engineering graduate with specialization coursework in NLP, Data Mining, and Pattern Recognition. Seeking senior or lead engineering roles at the intersection of AI infrastructure, cloud-native platforms, and distributed systems.

Cloud DevOps Engineer II — AI Platform R&D - Teradata - Remote

(2024-11)

• Architect and own a centralized AI platform standardizing LLM/ML access for 100+ enterprise users — managing GPU-aware scheduling (Nvidia), token rate limiting, semantic caching, tracing, and cost control via API Gateway (APISIX) and Istio Service Mesh across AWS, Azure, and GCP Kubernetes clusters (EKS, AKS, GKE, Rancher).
• Designed a production-grade multi-agent AI pipeline (LangChain, LangGraph, MCP, RAG, reasoning chains, MoE optimization, vector similarity, Pinecone/Postgres) saving 1,200–1,800 engineering hours annually ($90K–$135K USD) and reducing cross-team incident resolution by 97% — from 4 hours to under 30 minutes.
• Built a semantic caching layer over LLM API calls, cutting inference costs by 40% while maintaining response quality through embedding-based cache hit optimization.
• Architected enterprise-grade microservices for LLM training, Big Data, BI, and ML workloads using Spring Boot / Spring Cloud (Java 17), Go, and Rust, with hybrid Data Lake/Warehouse integrations and advanced API protocols (REST, gRPC, WebSockets, SSE, MCP).
• Implemented FIPS-compliant security controls with PKI, CMEK, TLS/mTLS, and NIST/RFC alignment; hardened IAM using Keycloak with LDAP, Active Directory, and OIDC for centralized SSO across 12+ applications, enforcing scalable RBAC/ABAC policies.
• Led cross-functional collaboration across UI/UX, Security, Infrastructure, and AI teams, reducing release friction by ~30% and streamlining integration across complex multi-environment pipelines.
• Created and led a structured onboarding program for 8 engineers, establishing coding best practices and a feedback system that decreased code review iterations by 30%.

Cloud Engineer II — Core Cloud Product - Teradata - Remote

(2022-11 - 2024-11)

• Designed and implemented cloud infrastructures across 5 environments (AWS, Azure, GCP, on-prem hybrid) using Docker, Kubernetes, and Python, ensuring high scalability and availability for Teradata's enterprise cloud product.
• Coordinated 10+ multi-team stories across 4 cross-functional teams (UI, Infrastructure, Network, Security), reducing multi-team issue resolution time from 6 to 4 weeks through architecture council participation and early edge-case identification.
• Standardized GitOps practices for 3 teams (~15 engineers working concurrently) by implementing pre-push hooks and fixing Jenkins pipeline reliability — reducing merge conflicts and CI failures by ~35%.
• Achieved 75–80% code coverage for continuous integration through rigorous unit testing (Python 3.6+), end-to-end testing, and extended logging module capabilities that reduced redundant data structures and improved test signal quality.
• Identified and eliminated unused cloud components, reducing team infrastructure costs by 10%, and collaborated with the Infrastructure team to recover costs from orphaned resources in failed deployments.

Software Engineer — Research Internship - Instituto de Ciencias Nucleares — UNAM - Remote

(2021-02 - 2021-10)

• Improved testability by +75% of original code coverage and reduced development cycle by 40% through codebase refactoring — applying design patterns and better programming techniques across Java and Python research software.
• Transformed team workflow using Git repository sharding with submodules and Git LFS, delivering 20% more features and resolving 30% more issues per sprint.
• Designed and maintained CI workflows with Docker and Jenkins, accelerating development cycles by 40%.
• Assessed feasibility of Java frameworks for scientific data processing and evaluated FreeRTOS as the main OS for embedded development boards used in the laboratory.

Freelance Software Developer - Self-Employed - Remote

(2017-01 - 2022-12)

• Delivered Java desktop and backend applications (Java 8/11/17, Spring Boot, Jakarta EE) for local businesses using MVC architecture, REST APIs, Hibernate/JDBC, and JavaFX/Swing GUIs.
• Built full-stack e-commerce platforms with the MEAN stack (Angular, Node.js, Express, MongoDB) using monorepo Git strategies, serving multiple SMB clients over 6 years.
• Developed Java EE web services with JSP/HTML5/CSS3/JavaScript frontend integrations and ORM-backed data layers, supporting 10+ client projects across diverse verticals.

Technical Support — Networking Internship - Instituto Politécnico Nacional (IPN) - Mexico City

(2016-11 - 2017-04)

• Configured and maintained 2 IPv4 networks and managed local server services (2 servers, 3 services) for on-premise institutional deployments.

B.Eng. - Computer Engineering - Facultad de Ingeniería — UNAM

Technical Degree - Computer Programming & Full Stack Java Development - Instituto Politécnico Nacional (IPN)