Multi-Agent Analytics for a Banking Lakehouse
Enterprise · Banking · In production
Multi-agent system (LangGraph + RAG over Azure Databricks) that automates metadata lineage exploration across a terabyte-scale Lakehouse at BCP.
Also known as Carlos Cesar Sanchez Coronel (Carlos César Sánchez Coronel).
AI Engineer · Data Engineer · AI Infrastructure & Operations
Agentic AI · Data Platforms · Production AI Ops
Industrial & Enterprise AI
Lima, Peru · UTC-5
Industrial & Enterprise AI
Open to remote & relocation · English B2 · Spanish native
Lead AI initiatives on the bank-wide Azure Databricks Lakehouse: multi-agent systems (LangGraph, RAG), terabyte-scale metadata engineering, AI governance with enterprise stakeholders.
Pricing & revenue-growth ML models (Microsoft Fabric), LLM-based automation for manual data processing, end-to-end PySpark pipelines across Azure, AWS, and on-prem.
Distributed telemetry pipelines (PySpark, Airflow, Kafka), metadata engineering, churn prediction models, and conversational SQL-agent prototypes over OpenAI APIs.
Built the company's first Data Warehouse (T-SQL, SSIS), Python RPA for legacy data extraction, and Power BI reporting for operations.
Techno-economic analytics for battery storage and power markets, predictive maintenance data, SCADA telemetry — the industrial foundation behind my AI work today.
M.Sc. in Artificial Intelligence, Universidad Nacional de Ingeniería. Thesis: Hybrid Edge-Cloud Routing for SLMs in Edge AI: Empirical Evaluation of Latency, Cost, and Resilience.
Part-time Lecturer of AI at UMA — Deep Learning, Machine Learning, and NLP. Open course tracks and workshops: