Hi, I'm
Murillo
Sezerino
Analytics Engineer
I build the transformation layer: from raw data to the dimensional model the business can actually use. I come from finance, credit and operations (Loggi, Stone, PagBank).
Data with business context.
Analytics Engineer · SQL · dbt · Python · dimensional modeling · data quality
Current focus: dimensional modeling, transformation with dbt, data testing and API delivery. I worked with portfolio risk, delinquency, delivery SLA and churn before I worked with pipelines, so I decide what the data needs to become before deciding how to move it.
Companies and clients I've worked with
Selected work.
ETL Pipeline · Delivery Logistics
Multi-source ETL in Python, from raw data to partitioned Parquet loads on object storage. Status normalization, UTC date parsing, Haversine distance calculation, a quality gate that stops bad data before the load and incremental watermark loading that is idempotent across runs.
Retail Data Platform
Analytical platform under construction: incremental ingestion, Airflow orchestration and dbt from staging to star-schema marts, with tests, docs and lineage, a warehouse and a dashboard. The retail analytics base is already in the repository, with RFM segmentation and churn prediction on 2,000 customers and 15,000 orders in PostgreSQL.
Credit Scoring · Stacking Ensemble
Default prediction with a calibrated LightGBM in serving and a Stacking Ensemble (XGBoost, LightGBM, CatBoost and Random Forest) in the study. SMOTE applied after scaling to avoid leakage, credit feature engineering, explainability with SHAP and delivery through a REST API in FastAPI.
Where I built this.
- I design and implement ETL pipelines in Python and SQL, automating ingestion, transformation and loading for clients across segments
- I build AI agents and automations with LLMs applied to data analysis, report generation and business processes
- I deliver end-to-end applications with Next.js and FastAPI, from the data pipeline to the final interface
- I implement analytical dashboards and real-time KPI systems, with serverless deploy and CI/CD via GitHub Actions
- Analyzed transactional behavior of a 180-client portfolio for profile segmentation and commercial prioritization
- Mapped churn causes from usage and cancellation data, and acted on measures to raise customer retention
- Delinquency and profitability dashboards in Power BI, used in the routine portfolio follow-up
- Raised SLA compliance through bottleneck analysis in the delivery flow, in-process orders and route restructuring
- Improved inventory accuracy with audit routines and continuous KPI monitoring
- Automated operational performance reports, consolidating indicators
- Automated ETL flows processing operations and logistics records, eliminated manual work and cut report generation by migrating data to SQL and Python
- Ran route optimization tests, with gains in delivery efficiency
- Led a team as a bridge between operations, business and technology
- Structured automated financial controls and KPI reports (delinquency, adoption, satisfaction) for the board
- Identified 8 delinquency risk factors to support changes in credit policies
- Corporate risk analysis of portfolios
- Automated the control of a 112-client portfolio, reducing processing time by 60%
Contact.
Open to Analytics Engineer / Data Analyst roles, remote or hybrid in the Vale do Paraíba region and São Paulo City.