Skip to content
Juan Sebastián Casallas

Juan Sebastián Casallas

Financial Data Analyst — Modeling & Automation

From data to decisions.

I build financial models that forecast Colombia’s usury rate within 2 basis points of the official figure and automate reporting pipelines from 6 hours to 15 minutes.

View projectsLearn more

Impact, in numbers.

  • 6 h → 15 min

    Reporting pipeline, automated end‑to‑end.

  • 2 bps

    Average error of the usury-rate model vs. the official figure.

  • ~95%

    Of the commercial portfolio covered by the repricing model.

About me.

I transform complex financial data into automated systems and predictive models that deliver measurable business results.

Industrial engineer turned financial data analyst.

Four years in banking, building the reporting pipelines and forecasting models that business teams rely on, with Python, SQL, R, and Power BI.

Open to financial data analyst roles where analytical thinking drives real business impact.

Let’s connect.

Work experience

  1. Financial Analyst II

    Banco de Bogotá

    Mar 2024 – Present

    • Automated a financial reporting pipeline with SQL and PL/SQL, cutting generation time from 6 hours to 15 minutes.
    • Rebuilt budget migration workflows with SQL and R, simplifying commercial management and reducing operational friction across teams.
    • Developed a repricing model for the commercial loan portfolio with SQL and R, covering ~95% of the portfolio with a maximum deviation of 1 basis point.
    • Built a macroeconomic variables observatory that consolidates the policy rate, IBR, DTF, CPI and UVR daily and projects them to 2030, feeding the repricing model.
    • Built an optimization model in R that estimates the usury rate (TIBC) over millions of records, within just 2 bp of the SFC.
    • Automated a data pipeline in R, cutting storage requirements by more than half while expanding historical coverage by 13 years.

  2. Pricing Analyst III

    Banco de Bogotá

    Sep 2022 – Mar 2024

    • Automated the extraction of banking tariffs from the Superfinanciera with Python web scraping, creating the sector’s first structured historical benchmark.
    • Developed a projection model for commissionable services in Excel and VBA, supporting pricing strategy with forecasts aligned to business goals.
    • Conducted market research and competitive analysis across 5 banking entities, identifying pricing trends and optimization opportunities in the sector.
    • Collaborated with multidisciplinary teams to design and monitor pricing strategies, ensuring profitability and market competitiveness.

  3. Data Analyst

    Numeris Consultoría Actuarial y Financiera

    Jul 2022 – Sep 2022

    • Cleaned and processed actuarial and financial insurance data in R, ensuring quality and consistency for decisions based on reliable information.

Core skills

  • Python
  • SQL
  • R

Data & analytics

  • pandas
  • NumPy
  • scikit-learn
  • DuckDB
  • Parquet
  • PL/SQL

Modeling

  • Financial modeling
  • Pricing
  • Forecasting
  • Optimization
  • Regulatory analytics
  • Machine learning

BI & automation

  • Power BI
  • Power Automate
  • Excel & VBA
  • ETL
  • Web scraping
  • Git

AI-assisted development

  • Claude
  • Gemini
  • LLM automation

LanguagesSpanish (native) · English (B1)

Recognitions

Most Innovative

Strategy & Profitability Division

Aug 2023

Awarded first place in an internal innovation contest organized by the Strategy and Profitability Division for developing an automated Python tool that extracted and processed historical tariff data from the Superfinanciera. It gave the division its first structured historical benchmark of banking system tariffs.

Education

Industrial Engineering

Pontificia Universidad Javeriana

Jan 2017 – Jun 2022

Degree thesis: Design of a demand forecasting model based on machine learning and a multi-objective model for production planning in a bakery industry

Read thesis

Certifications

  • Machine Learning Specialization

    DeepLearning.AI & Stanford Online

    Mar 2025

  • Data Analysis using R

    Google

    Feb 2025

  • Effective Problem Solving

    University of California, Irvine

    Feb 2025

  • Data Storytelling Practices

    Crehana

    Jan 2024

Case studies.

Select projects in advanced analytics, automation, and predictive modeling.

  • Python
  • DuckDB
  • Regulatory Modeling
  • Financial Analytics

Usury Rate Forecasting Model (TIBC)

Current usury rate · Oct 202628.59%TIBC 19.06 % · −65 bp vs. previous month

2 bp average error vs. the official SFC rate

A regulatory simulation model that anticipates Colombia’s maximum lending rate week by week, enabling the finance team to forecast revenue impact before the official SFC publication.

View demo
  • Python
  • pandas
  • GitHub Actions
  • Macroeconomic Modeling

Macroeconomic Variables Observatory

Live data · updated every Monday

Policy rate12.25%+25 bp in 30 days

7 daily series projected to 2030; the UVR matches the official figure

Consolidates Colombia’s policy rate, IBR, DTF, inflation and UVR day by day and projects them to 2030 with Banco de Bogotá’s economic research scenario.

View demo
  • R
  • Power BI
  • Macroeconomic Modeling
  • Financial Analytics

Commercial Portfolio Repricing Model

~95% of the commercial portfolio, within 1 bp of the system’s daily rate

A rate-sensitivity model that estimates the daily and monthly impact of macroeconomic variable changes on the commercial lending portfolio, enabling structured weekly revenue planning for the largest segment of the bank’s commercial loan book.

View demo
  • Academic
  • Python
  • scikit-learn
  • ML
  • Optimization

Industrial Bakery Demand ML Model

An end‑to‑end machine learning and optimization system that forecasts demand and plans production for an industrial bakery, optimizing for both cost and service level.

Read thesis

Let’s build something.

Get in touch for opportunities, collaborations, or just to say hi.

Contactos_JS@pm.me