Cloud Solutions Architect

Designing scalable systems, integrating modern data pipelines, and applying AI-driven solutions to solve complex business challenges.

I build marketing automation, cloud architectures, and data integrations that serve real-world outcomes. Always learning, always improving.

View Projects
Mauricio Guzman Headshot

Core Capabilities

Cloud & Architecture

AWS · Azure · Serverless · API design · Identity · Security · Scalability

Data & Integrations

Snowflake · ETL · SQL · Python · Data modeling · Pipelines · Streaming

MarTech & Automation

Salesforce Marketing Cloud · Journey design · AMPscript · Automation Studio

AI & Emerging Tech

LLM workflows · Agents · AI-assisted architecture · Prompt engineering

Development

Python · JavaScript · Flask · REST APIs · Web development

Consulting & Strategy

Blueprinting · Technical discovery · Solutions alignment · Stakeholder guidance

Experience

Salesforce Marketing Cloud Solutions Architect
2023 – Present

Lithia & Driveway · Remote

  • Enterprise Cost Savings & Product Delivery: Led the architectural design and rollout of an internal Reputation Management system, replacing third-party vendors to save the company ~$1.8M annually. Coordinated with Product and Executive leadership to meet a strict 8-month go-live deadline.
  • Migration & Scale: Managed the technical migration of over 200 dealership stores on launch day, consolidating disparate local Business Unit (BU) journeys into a centralized architecture to streamline operations.
  • Technical Solutioning: Designed the backend data flow using complex SQL automations and secure FTP integrations to sync feedback data with external partners. Validated the system through pilot testing and led the final "switch-over" strategy.
Marketing Data Engineer
2022 – 2023

Lithia & Driveway · Remote

  • Analytics Infrastructure Pilot: Engineered a Google BigQuery SQL solution to solve a reporting gap for the Creative Team, linking GA360 data (across 300+ store views) with Adobe Workfront asset logs via FTP. This enabled the first-ever tracking of asset-specific performance metrics in Power BI.
  • Proof of Concept to Enterprise Strategy: Successfully deployed the scheduled query automation to prove value. Identified long-term scalability and cost constraints (Power BI timeouts/BigQuery costs) and used these findings to guide the Data Team toward a robust, permanent Snowflake implementation.
  • Pipeline Maintenance: maintained reporting pipelines and collaborated with engineering to define data governance standards for downstream dashboards.
Software Quality Assurance Lead
2015 – 2022

Transplant Connect · Hybrid

  • HIPAA Compliance & Reliability: Owned the testing strategy for mission-critical organ donation software. Ensured 100% compliance with strict healthcare regulations before any code hit production.
  • Automation Strategy: Directed the scope of automated testing by identifying high-risk workflows for the automation team to target. Personally contributed to script development and maintenance to reduce manual regression cycles.
  • Risk Assessment: Acted as the final "gatekeeper" for releases, proactively identifying architectural risks and scalability bottlenecks to prevent post-deployment failures.

Projects & Case Studies

A mix of cloud, data, automation, AI, and integration work.

AI-Assisted Cloud Architecture Intake System

Next.js OpenAI GPT-4o-mini Vercel Serverless Functions
Live Demo GitHub
Technical Architecture
Frontend & State Management

Built on Next.js to utilize a modular, component-based architecture mirroring enterprise standards. Leveraged Serverless API Routes to secure API keys and orchestrate backend requests.

AI Model Selection

Selected GPT-4o-mini as the inference engine. This offered the optimal intersection of latency, cost-efficiency, and reasoning quality for high-level architectural abstractions.

Deployment Strategy

Deployed via Vercel to utilize edge caching and automated CI/CD pipelines, ensuring Day 1 scalability.

Key Decisions & Trade-offs
Probabilistic vs. Deterministic

Constraint As an LLM-based solution, outputs are probabilistic.
Mitigation UI explicitly frames results as "Recommendations" to set user expectations for discovery, not final implementation.

Domain Specialization

Leverages a generalist foundation model. Effective for standard patterns (e.g., 3-Tier Web App) but lacks deep specialization for niche compliance (HIPAA).

External Dependency

Architectural Note: Tightly coupled to OpenAI API limits. A production-grade version would require a fallback strategy (e.g., queuing requests) to handle downtime.

RESTful Vehicle Data API
GitHub

Python · Flask · API Design

Designed a lightweight REST API to decouple vehicle data logic from frontend applications. Implemented standardized JSON response schemas and error handling patterns suitable for microservices consumption.

Automated Weather Data ETL
GitHub

Python · SQLite · Plotly · Automation

Engineered an end-to-end data pipeline ingesting real-time API data. Automated the Extraction, Transformation, and Loading (ETL) into a relational store (SQLite) to power dynamic visualization dashboards.

Get in Touch

Open to consulting, contract work, and architecture roles.

Email: