𝐋𝐞𝐝 𝐭𝐡𝐞 𝐞𝐧𝐝-𝐭𝐨-𝐞𝐧𝐝 𝐝𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭 of a set of credit scoring machine learning models for Personal Loan and Buy Now, Pay Later (BNPL) products to 𝐩𝐫𝐞𝐝𝐢𝐜𝐭 𝐝𝐞𝐟𝐚𝐮𝐥𝐭 𝐫𝐢𝐬𝐤.• Managed key processes including EDA, model evaluation, hyperparameter tuning, deployment, and monitoring.• Included Product-Specific Models, Behavioral Models, and a mix of both.• 𝐈𝐧𝐜𝐫𝐞𝐚𝐬𝐞𝐝 𝐥𝐞𝐚𝐝 𝐠𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐨𝐧 𝐛𝐲 𝐨𝐯𝐞𝐫 𝟓𝟎% while maintaining the same risk appetite.𝐃𝐞𝐬𝐢𝐠𝐧𝐞𝐝, 𝐝𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐝, 𝐚𝐧𝐝 𝐢𝐦𝐩𝐥𝐞𝐦𝐞𝐧𝐭𝐞𝐝 an income prediction machine learning model to 𝐞𝐬𝐭𝐢𝐦𝐚𝐭𝐞 𝐧𝐞𝐭 𝐬𝐚𝐥𝐚𝐫𝐢𝐞𝐬.• 𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐝 𝟓𝟎% 𝐩𝐫𝐞𝐜𝐢𝐬𝐢𝐨𝐧, significantly improving the accuracy of risk assessment measures.𝐃𝐞𝐬𝐢𝐠𝐧𝐞𝐝 𝐚𝐧𝐝 𝐢𝐦𝐩𝐥𝐞𝐦𝐞𝐧𝐭𝐞𝐝 𝐚 𝐩𝐫𝐨𝐩𝐞𝐧𝐬𝐢𝐭𝐲 𝐦𝐨𝐝𝐞𝐥 to predict the likelihood of leads adopting new financial products.• Improved marketing precision and acquisition rates

𝐋𝐞𝐝 𝐭𝐡𝐞 𝐚𝐮𝐭𝐨𝐦𝐚𝐭𝐢𝐨𝐧 of teacher recruitment operational processes through AI.• 𝐑𝐞𝐝𝐮𝐜𝐞𝐝 𝐂𝐕 𝐫𝐞𝐯𝐢𝐞𝐰 𝐭𝐢𝐦𝐞 𝐛𝐲 𝐨𝐯𝐞𝐫 𝟒𝟎%.𝐂𝐨𝐧𝐝𝐮𝐜𝐭𝐞𝐝 𝐰𝐞𝐛 𝐬𝐜𝐫𝐚𝐩𝐢𝐧𝐠 on portals such as Laborum, Bumeran, Computrabajo, and LinkedIn, applying 𝐭𝐞𝐱𝐭 𝐚𝐧𝐚𝐥𝐲𝐭𝐢𝐜𝐬 and 𝐩𝐫𝐨𝐦𝐩𝐭 𝐞𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 (𝐎𝐩𝐞𝐧𝐀𝐈) to improve the educational offering.• Enabled the deployment of 𝟔 𝐭𝐨 𝟖 𝐧𝐞𝐰 𝐩𝐫𝐨𝐠𝐫𝐚𝐦𝐬 𝐩𝐞𝐫 𝐬𝐞𝐦𝐞𝐬𝐭𝐞𝐫.

𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐝 𝐚𝐧𝐝 𝐥𝐞𝐝 𝐭𝐡𝐞 𝐞𝐧𝐝-𝐭𝐨-𝐞𝐧𝐝 creation of 𝐬𝐢𝐱 𝐜𝐡𝐮𝐫𝐧 𝐩𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 𝐦𝐨𝐝𝐞𝐥𝐬 per semester, supervising the entire client life cycle.• Covered key areas: exploratory data analysis (EDA), model evaluation, model tuning, model explainability analysis, and result visualization.• 𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐝 𝐜𝐨𝐯𝐞𝐫𝐚𝐠𝐞 𝐚𝐧𝐝 𝐩𝐫𝐢𝐨𝐫𝐢𝐭𝐢𝐳𝐚𝐭𝐢𝐨𝐧 𝐟𝐨𝐫 𝐨𝐯𝐞𝐫 𝟖𝟎% 𝐨𝐟 𝐝𝐞𝐬𝐞𝐫𝐭𝐞𝐫𝐬𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬: Python, Pyspark, Scikit-learn, XGBoost, LightGBM, CatBoost (ML), Seaborn, Matplotlib (visualization), SHAP (model explainability).𝐄𝐱𝐞𝐜𝐮𝐭𝐞𝐝 𝐚𝐮𝐝𝐢𝐨 𝐭𝐫𝐚𝐧𝐬𝐜𝐫𝐢𝐩𝐭𝐢𝐨𝐧𝐬 and 𝐭𝐞𝐱𝐭 𝐚𝐧𝐚𝐥𝐲𝐭𝐢𝐜𝐬 to identify improvement opportunities in sales processes.• Refined sales agents' speech with differentiating elements and a standard structure.𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬: Python, Whisper (audio transcription), NLTK (natural language toolkit), Stanza (natural language analysis).𝐃𝐢𝐬𝐜𝐨𝐯𝐞𝐫𝐞𝐝 𝐡𝐢𝐝𝐝𝐞𝐧 𝐭𝐨𝐩𝐢𝐜𝐬 in job offer documents.• Performed web scraping (Laborum, Bumeran, Computrabajo & LinkedIn), document cleaning and preprocessing, applied text topic modeling, and visualized results.𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬: Python, Selenium (web scraping), Gensim (topic modeling), NLTK (natural language toolkit), Stanza (natural language analysis).

Developed exploratory data analysis (EDA) and designed features for a predictive model, evaluating the impact of teacher training hours on teaching promotion evaluations

• Implemented customer segmentation using RFM analysis to identify outstanding and at-risk customers• Created commercial management and sales funnel reports with Python and SQL, integrated with Power BI and Looker Studio𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲𝘀: Python, PostgreSQL, Power BI

Built an application for data extraction and structuring through web scraping and API consumption, with interactive graph generation. This tool reduced report creation time by more than 50%

• Extraction, analysis and management of data• Creation of reports and dashboards• Support the automation of the unit's processes𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲𝘀: Python, Google Apps Script, Looker Studio

• Accidents' process mapping in SCTR Insurance to propose improvement opportunities• Development of cost reduction models in SCTR Insurance• Generation of insurance affiliation, renewal and disaffiliation information• Support in process improvement projects using agile methodologies𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲𝘀: MS SQL Server, Python, MS Excel