Portfolio

Selected work by our engineers across AI, full-stack, DevOps, and data projects. Each case study notes where the work was done.

Sports AnalyticsBusiness IntelligenceMachine LearningPredictive AnalyticsFEMEXFUTSports Intelligence

Sports Analytics Infrastructure & Match Performance ML Models

Founded Sports Business Analytics (SBA) and directed Sports Intelligence consulting projects for 5+ professional sports entities, including a flagship engagement with the Mexican Soccer Federation (FEMEXFUT). Built central Business Intelligence systems and deployed machine learning models to analyze player performance metrics, match statistics, and operational business data.

Organizational AnalyticsHuman Resources TechPsychometricsLeadership ConsultingExecutive CoachingWorkplace Culture

Data-Driven Organizational Climate & Leadership Development

Led organizational research and culture consulting initiatives at SOPHIA for corporate clients with workforces exceeding 800 employees. Designed psychometric and mixed-methods evaluation tools to analyze workplace burnout, stress patterns, and employee engagement. Developed humanistic leadership curricula, executive coaching frameworks, and cultural transformation strategies backed by empirical statistical analysis.

Workflow AutomationAI Customer ServiceProcess OptimizationHealthcare TechData PipelinesClinical Systems

Healthcare Operations & AI Customer Service Automation

Managed technology and operational transformation at Dialéctico (DBT Research Institute and Clinic). Designed and deployed AI-powered customer service systems to optimize patient intake and query handling. Automated internal administrative and research processes—streamlining data collection, psychometric reporting, and clinical quality compliance.

Reinforcement LearningQ-LearningDeep Q-Networks (DQN)PythonMATLABApplied AIUrban PlanningAcademic Research

Autonomous Urban Design via Deep Reinforcement Learning

Directed the "Urbanist RL" research group at Universidad Panamericana, leading 7+ researchers in developing autonomous AI agents for urban planning and resource allocation. Applied Q-Learning and Deep Q-Network (DQN) architectures to model urban layouts and labor market projections under game theory frameworks, resulting in 4+ peer-reviewed indexed publications and international conference presentations (MICAI, COMIA).

Process AutomationPower AutomatePower BICloud IntegrationWorkflow OptimizationData PipelinesSystem Integration

Cross-Departmental Cloud Automation & BI Integration

Executed full-cycle automation and data integration projects for commercial clients at Algoritso AI. Streamlined interdepartmental data flows by engineering cloud storage pipelines and automated Power Automate routines. Resolved complex system integration issues (including SMTP gas-industry messaging channels) and unified disparate data streams into central executive Power BI dashboards.

Data Science ConsultingMachine LearningAdvanced AnalyticsBusiness IntelligencePythonPredictive ModelingIndustrial AI

Industrial Predictive Analytics & Custom ML Solution Delivery

Spearheaded AI and Data Science consulting engagements for 5+ industrial enterprises under Algoritso AI. Developed specialized machine learning algorithms and predictive analytics models to improve operational efficiency, supply chain tracking, and executive decision-making. Managed a team of 7+ consultants to deliver custom enterprise data strategies and corporate AI training.

Frontend DevelopmentReactNext.jsJavaScriptUI/UX DesignREST APIsWeb ApplicationsProperty Management AI

Frontend Architecture for AI Property Management Platform

Led frontend development for landlord.ai, engineering the web user interface for property owners and managers. Directed UI/UX design workflows and integrated responsive React and Next.js interfaces with AI backend services and REST APIs, delivering an interactive dashboard for property management, tenant tracking, and property analytics.

Chief AI OfficerData SciencePredictive ModelingChurn AnalysisCustomer SegmentationPower BIPythonBusiness IntelligenceData Governance

Enterprise AI, Data Science & Business Intelligence Department Buildout

Served as Chief AI Officer for LUYOA, establishing the organization’s foundational Data Science, Business Intelligence, and AI function. Integrated enterprise databases with Jupyter notebook environments and Power BI to build automated end-to-end analytics pipelines. Led a team of 5+ data specialists in deploying supervised and unsupervised machine learning models for customer segmentation, behavioral forecasting, and churn prediction while establishing cross-departmental KPI dashboards for executive leadership.

Generative AIAWS BedrockAnthropic ClaudeAWS LambdaAPI GatewayAmazon S3PythonPrompt EngineeringServerless Architecture

First Responder Emergency Chat AI Summarization POC

Designed and executed a serverless AI proof-of-concept for Mobility4PS / Bridge4PS (via A2C Consulting) to automate intelligence reporting from first responder crisis chat channels. Engineered prompt-engineering frameworks on AWS Bedrock using Anthropic Claude models to parse unstructured chat histories and extract key entities (suspect names, locations, case numbers). Integrated AWS Lambda, Amazon API Gateway, and S3 to automatically merge AI-generated narratives into department-specific standardized report templates.

Generative AIAWS BedrockPythonFastAPIDockerAmazon ECSMicroservicesCI/CDGitHub ActionsPostgreSQL RDSMLOps

Enterprise Generative AI Microservice & AWS Bedrock Migration

Architected and delivered a standalone AI summarization and feedback microservice for EdTech platform ASSISTments (via A2C Consulting). Extracted complex LLM logic out of a legacy Java Spring monolith into a lightweight Python (FastAPI) application deployed via Amazon ECS/EC2 and ECR. Migrated the underlying LLM integration from Google Gemini to AWS Bedrock, configured Bedrock Guardrails for safety, provisioned dedicated PostgreSQL RDS instances, and built automated CI/CD pipelines via GitHub Actions with multi-model evaluation baselines.

Machine LearningAWS SageMakerPythonscikit-learnPostgreSQLSnowflakeTerraformCloudWatchMLOpsAgile PM

Scalable ML Constituent Scoring Pipeline Migration

Technical migration of an ML constituent scoring pipeline for non-profit fundraising platform GiveCampus (via A2C Consulting). Eliminated infrastructure bottlenecks caused by in-memory processing of tens of millions of records by orchestrating serverless AWS SageMaker Processing, Training, and Batch Transform jobs over PostgreSQL and Snowflake sources. Decoupled real-time application triggers from heavy inference workloads and implemented Infrastructure-as-Code (Terraform) alongside CloudWatch monitoring and automated error handling under AWS Well-Architected guidelines.

Cloud MigrationGoogle Cloud PlatformAWSBigQueryRedshiftdbtLookMLLookerTableauData EngineeringSQL

Cloud Analytics & Reporting Infrastructure Modernization

Directed the complete migration of core enterprise reporting infrastructure for online retail platform Custom Ink. Engineered scalable data models in LookML layered over dbt staging models and BigQuery/Redshift data warehouses. Configured secure server-to-server integration points, automated CI/CD deployment workflows, and transitioned SAP business intelligence feeds into real-time Looker dashboards for finance, marketing, customer service, and operations stakeholders.

AI EngineeringFull StackNext.jsMCPAWS

Iterca Coaching Platform

Built for Iterca as platform engineer. Next.js 15 (App Router, React Server Components) on AWS Amplify, with PostgreSQL on RDS inside a private VPC. A double-entry financial ledger with strict concurrency controls keeps credits from being double-spent, even under scripted multi-tab checkout attempts, and a webhook-driven payout state machine (Calendly + Payoneer) handles payouts for 50+ coaches. A standalone Model Context Protocol server gives AI assistants role-based access to meeting notes without exposing private data.

AI EngineeringRAGLLMn8nPostgreSQL

SigenStor AI Agent

A multi-turn sales agent with session state in PostgreSQL and a vector-embedding RAG pipeline over technical manuals and pricing sheets, so every answer traces back to source text. Battery sizing pulls live 3-phase load data from HomeWizard telemetry and checks it against hard installation constraints (cable cross-section, floor loading), so it only recommends installs that can actually be built.

AI EngineeringOCRLLMPython

Document Intelligence & Form Automation

Replaced manual transcription with Google Cloud Vision OCR and OpenAI entity and KPI extraction, feeding a dual-view review screen: the scanned pamphlet on one side, pre-filled fields on the other. Operators confirm and submit instead of retyping, which cut per-record handling time dramatically across a backlog of 50k+ pamphlets.

Full StackNode.jsKafkaMongoDBElasticsearch

Lead & Booking Platform (TRA)

Co-architected the TRA API (Node.js/TypeScript) and the TRADB MongoDB schema shared by Trident's booking and sales apps. A message broker absorbs high-volume call traffic, WebSockets push live reservation updates, and Elasticsearch indices stay in sync for search. Household and lead merging logic was covered by test suites for 20+ edge cases before reaching production, and re-tuning the MongoDB cluster saved about $12,000/year with no performance loss.

Full StackMobileFlutterRedisMongoDB

iRide Ride-Hailing App

Built the iRide app for iRide PK in Flutter with BLoC state management. As usage grew, Redis caching and MongoDB query tuning cut hosting costs by 80%, from $1,500 to $300 per month.

AI EngineeringSpeech-to-TextWebSockets

Voice Intelligence & Real-Time Transcription

A WebSocket microservice streams live call audio to IBM Watson Speech-to-Text, so transcripts keep pace with the call instead of being batch-processed afterward. A rule layer on the transcript stream flags moments where a caller asks for passcodes or restricted data, giving compliance teams a short list of calls worth reviewing.

DevOpsKubernetesCI/CDKafkaJava

Cloud Infrastructure & Event-Driven Services

Led Kubernetes-based deployments and CI/CD improvements (Jenkins, GitHub Actions) with infrastructure as code at Trident Marketing. Designed SWAM, a Quarkus 3 incident platform in hexagonal architecture: a Kafka event triggers a separate analyzer service for Postgres trigram duplicate detection, so the main API never waits on it. Shipped with Docker Compose locally and Kubernetes manifests in the cluster.

Data Science & MLHealthcareFlutterVue.js

HCAI: Post-Surgical Infection Tracking

Clinicians log LAARC-standard lab evaluations from the ward in a Flutter app, feeding a Vue.js dashboard that WHO and NIH collaborators use for performance reporting. As solutions engineer, designed the microservice split between mobile intake and reporting, and a data model built around the surveillance protocols the team needed.