Sergio Echeverri

AI/ML Engineer with 8+ years of experience designing and deploying production-grade machine learning and LLMsystems for fintech and insurance platforms. Specialized in ML pipelines, NLP/LLM applications, scalable MLOpsinfrastructure, and data engineering workflows that power intelligent automation, risk prediction, and decision-support services.Professional ExperienceSenior Software Engineer, Allio Capital• Designed and implemented end-to-end ML pipelines using Python, Scikit-learn,PyTorch, and TensorFlow for investor behavior modeling, churn prediction, portfoliooptimization, and risk scoring.• Built feature engineering and dataset preparation frameworks using Spark, Pandas,Airflow, and SQL data warehouses, enabling reproducible and scalable trainingworkflows.• Developed LLM-powered financial research assistants leveraging embeddings,vector databases, and Retrieval-Augmented Generation (RAG) to answer analyst andinvestor queries over proprietary financial data.• Implemented real-time inference services using FastAPI, Docker, and Kubernetes,serving prediction and recommendation APIs integrated into production tradingsystems.• Established full MLOps infrastructure with MLflow model registry, automatedretraining pipelines, drift monitoring, and CI/ CD deployment workflows on AWSSageMaker and EKS.• Integrated ML outputs into portfolio advisory, compliance, and risk managementservices via REST APIs and Kafka event streams.• Built model monitoring dashboards tracking latency, accuracy, and data drift usingPrometheus and Grafana.Software Engineer, Goldman Sachs• Built large-scale data processing pipelines using Python, Spark, SQL Server, andKafka to support ML model training and enterprise analytics workloads.• Developed experiment tracking and model evaluation platforms for internal MLresearch teams, enabling systematic comparison of model performance.• Integrated trained model inference outputs into pricing, risk, and trading platformsthrough scalable API services.• Implemented data validation, lineage tracking, and quality monitoring to ensureintegrity of ML training datasets.• Collaborated with quantitative and ML engineers to transition research models intoproduction systems.