Cristian Correa Sandoval

AI Engineer · Data Scientist

Industrial Engineer, Universidad de Chile  ·  Santiago, Chile
ccorrea.s@hotmail.com  ·  +56 9 6320 5715

Download CV (PDF) LinkedIn

Summary

AI Engineer and Industrial Engineer (Universidad de Chile) with 6+ years turning data into business decisions — from ETLs and optimization models to ML in production and, currently, generative-AI conversational agents. I build LLM-based agents end-to-end: LangGraph orchestration, FastAPI services, distributed memory, end-to-end infrastructure on GCP via Terraform, and node-level observability with Langfuse.

Experience

AI Engineer Falabella Retail · Santiago, Chile 2025 – Present
  • Sales conversational agent: drives product sales with ~10s average latency per message, backed by a two-tier memory design — short-term in Redis, long-term in Firestore and PostgreSQL — for persistent, multi-session context.
  • Business-insights agent (Looker Studio): generates dynamic SQL queries on demand and returns tailored business reports with ~1 min average query latency, plus error-recovery and governance mechanisms.
  • Infrastructure & CI/CD: built the full CI/CD pipeline with Terraform (IaC), provisioning the complete GCP stack (Cloud Run, Firestore, Redis, Cloud Storage, Secret Manager).
  • Observability: instrumented node-level tracing of agent graphs with Langfuse to monitor cost, latency, and quality in production.
Data Scientist Consorcio · Santiago, Chile 2023 – 2025
  • Led end-to-end data science initiatives with ML Engineers, from label engineering and EDA to deploying and monitoring models in production on AWS.
  • Propensity models (banking & insurance): XGBoost regression for lead propensity + multilabel classification for product recommendations, improving targeted marketing.
  • Contactability optimization: predictive model prioritizing the best phone numbers using internal and purchased data; raised reachability and lowered acquisition cost via uncertainty-based optimization.
  • Scalable ML pipelines on AWS Sagemaker: distributed training, async and multi-model endpoints, Athena integration, and compressed-JSONLines batch inference, reducing inference time and cost.
  • Generalized pipeline architecture across projects and ran internal upskilling workshops.
Optimization Projects Leader SCM Latam · Santiago, Chile 2023
  • Led development of products with advanced optimization engines and consolidated the area.
  • Built engines for optimal staffing per contract, overtime offsetting, regulation-compliant work-schedule generation, and efficient schedule allocation.
Workforce Management Consultant SCM Latam · Santiago, Chile 2021 – 2022
  • Developed Python optimization models for HR challenges across multiple countries and sectors.
  • Led implementation of tech solutions and creation/enhancement of cloud-based products.
Thesis Student GlobeItalia · Santiago, Chile 2021
  • Researched and implemented optimization models to forecast monthly product demand and reduce stockouts, automating result generation with Python.
Data Analyst Falabella Financiero · Santiago, Chile 2020 – 2021
  • Built Tableau dashboards on financial indicators and automated their refresh with Python ETLs.
Business Process Analyst Seguros Falabella · Santiago, Chile 2019 – 2020
  • Mapped company processes (As-Is flows) and ran activity studies to generate efficiencies; led improvement of the paperless project.

Technical Skills

Languages
PythonSQLRJuliaLua
GenAI / LLMs
LangGraphGoogle ADKLangfuseRAGAgent design
ML / DS
XGBoostscikit-learnLabel engineeringBatch inferenceBatch trainingFine-tuningModel evaluation
Backend
FastAPIPostgreSQL
Cloud / MLOps
AWS (Sagemaker, Athena, S3, RDS)GCP (Cloud Run, Firestore, Redis, Secret Manager)DockerTerraformCI/CD
BI / Viz
Looker StudioPower BITableauD3.js

Education

Universidad de Chile 2014 – 2022 Ingeniería Civil Industrial (Industrial Engineering) · Santiago, Chile