Germán Goñi Núñez

Engineer with 12+ years of Python experience building production-grade AI agentic systems. Currently architecting multi-agent orchestration with LangGraph and ADK at Latam Airlines, processing real-time telemetry into predictive maintenance recommendations — on track for $35M USD in operational savings. Built RAG pipelines with semantic chunking, vector retrieval, and source-traceable responses for technical documentation.

Designed and deployed AI agent integrations with enterprise APIs, databases, and third-party services across 5+ industries. Backend-first engineer who treats AI as a layer on top of solid engineering fundamentals: CI/CD, testing, monitoring, and clean architecture.

Digital Team Lead: AI & Data Monetization at Latam Airlines (2024-05 – Present)

Architecting multi-agent orchestration with LangGraph and ADK, managing AI lifecycle from prototype to production with MLOps discipline

• Multi-agent orchestration with LangGraph + ADK: Designed and deployed specialized agents for data ingestion, causal inference, and alerting — integrated with enterprise APIs for predictive maintenance, on track for $35M USD savings in 2025
• RAG pipeline for technical documentation: Built retrieval-augmented generation system processing aircraft maintenance manuals with semantic chunking, vector retrieval, reranking, and source-traceable responses — enabling context-aware agent decisions
• Backend Python services: Built end-to-end data pipelines: ingestion → feature engineering → model inference → API exposure, with CI/CD, monitoring, and production-grade reliability
• Managed full AI lifecycle from prototype to production with rigorous MLOps discipline, collaborating across data science, data engineering, and product teams

Lead Consultant & Partner, AI/ML Solutions at Anticiparte (2019-12 – Present)

Enterprise AI/ML consulting focused on API integration, data pipelines, and computer vision deployment

• Enterprise API integration (Banco de Chile): Built ML-driven risk models integrated with contact center systems via REST APIs, reducing costs by 60% and improving debt recovery by 10%
• Data pipeline engineering (Sodexo): Engineered time-series forecasting pipeline achieving 5% MAPE, reducing processing time from 7 hours to 10 minutes — exposed predictions via internal APIs to operational systems
• Computer Vision at scale (Pacific Mining Parts): Deployed CV/ML system at Chuquicamata reducing false positive alarms by 85% (20% → 3%), integrated with plant monitoring infrastructure

Operations Lead, Analytics & Risk Modeling at Fintonic (2018-10 – 2019-11)

Led analytics and risk modeling operations for fintech lending platform

• Built ML-driven risk models for credit lending, optimizing interest rate spreads and default mitigation
• Hit 200K active users target ahead of schedule using Advanced Analytics for conversion funnel optimization
• Secured B2B integrations with banks and local insurers, establishing new revenue channels

Leadership Roles: Digital Transformation at ACHS / Cencosud (2012 – 2018)

Led digital transformation initiatives across healthcare and retail divisions

• Implemented biometric access/time-tracking across 70 clinical centers; orchestrated Azure analytics initiative for payroll automation
• Led budget control for 5,000+ employees at Cencosud retail division

MSc in Artificial Intelligence – Universidad Adolfo Ibáñez

B.Sc. in Industrial Engineering (Mining Minor) – Pontificia Universidad Católica de Chile

Program Certificate in Digital Transformation – MIT / Emeritus Institute