Ignacio Espejo

Background

About Ignacio Espejo

I build production ML systems that deliver measurable business impact. I've architected and deployed 18+ serverless ML pipelines processing 100M+ records monthly, generating CLP CLP in quantified business value through: • Revenue Protection:$220M+ annually through demand forecasting systems (MAPE <5%) preventing fuel stock depletion across 600+ service stations • Cost Optimization:$590M+ in electricity savings via predictive models for contract negotiation and automated tariff optimization • AI Automation: 90%+ reduction in manual effort through multi-LLM ensemble systems (GPT-4, Claude) processing 100+ contracts daily • MLOps Excellence: Accelerated deployment cycles from months to weeks with generalized Infrastructure as Code framework MY APPROACH TO ML ENGINEERING: → Business Impact First: Every model must drive quantifiable value → Production-Ready: Cloud-native, scalable, monitored systems from day one → Full-Stack ML: From data pipelines to model deployment to stakeholder collaboration → Modern Stack: AWS (CDK, SageMaker, Lambda, Step Functions), LLMs, MLOps best practices TECHNICAL EXPERTISE: • Cloud-Native MLOps: AWS ecosystem (CDK, SageMaker, Glue, Lambda, Fargate, Step Functions) • Time Series Forecasting: Prophet, NeuralProphet, SARIMAX, Holt-Winters (MAPE 1.23%-5%) • LLM Engineering: GPT-4, Claude 4/4.5, AWS Bedrock, prompt engineering, ensemble architectures • Production ML: 18+ deployed pipelines, 100M+ records/month, automated monitoring • Infrastructure as Code: AWS CDK, multi-environment CI/CD, reusable patterns Santiago, Chile | Open to remote opportunities gn.espejo@gmail.com