Aaron ❅ Castillo Email & Phone Number

Mexico City, Mexico

Mexico City, Mexico

TECH STACK: Python, Java, Redshift, MySQL, Talend (TAC, TMC), Hashicorp Vault, Airflow, FastAPI, AWS (S3, SQS), Google (Forms, Colab), Power BI, Coderbytes, Google Apps Scripts, Jira, OS TicketWITH THE CLIENT (Arcade/Entertainment Company):• Successfully implemented Service Level Agreement policy, along with its proper installation in both OS Ticket and Jira for a 500-employee company.WITH THE CLIENT (Pharmaceutical Company):• 20% time reduction on ETL processes… Show more TECH STACK: Python, Java, Redshift, MySQL, Talend (TAC, TMC), Hashicorp Vault, Airflow, FastAPI, AWS (S3, SQS), Google (Forms, Colab), Power BI, Coderbytes, Google Apps Scripts, Jira, OS TicketWITH THE CLIENT (Arcade/Entertainment Company):• Successfully implemented Service Level Agreement policy, along with its proper installation in both OS Ticket and Jira for a 500-employee company.WITH THE CLIENT (Pharmaceutical Company):• 20% time reduction on ETL processes with Python scripts and AWS (S3, Redshift) for 500GB of data on a regular basis.• 25% increase in delivery speed on reporting topics by using Airflow, Python and Power BI.• 25% increase in learning speed regarding onboarding topics by creating time-varying documentation with Javascript and HTML.AT SVITLA:• 25% successful hires as part of my Data interviewer role.• Executed a LSTM neural network for Sentiment Analysis evaluation on customer reviews and corporate assessments, with outstanding results.• As an Application Support Lead, I gained excellent feedback by proposing efficient and disruptive methodologies for tech apprenticeship. Show less

Mexico City, Mexico

TECH STACK: Python, Scala, PySpark, Javascript, Ubuntu, Couchbase, Bash, AWS(Athena, Glue, Cloudwatch, S3, SQS) Google (Apps Scripts, Forms), Zapier, YouCanBookMe.WITH THE CLIENT (Online Privacy Company):• 460% time improvement (from 23 hours to 5) by designing and developing a data pipeline in AWS Glue and PySpark for 1 TB of data processing every day.• 50% time reduction (from 8 to 4 hours) on bulk data transfer activities by emulating AWS functionality (aws s3 cp… Show more TECH STACK: Python, Scala, PySpark, Javascript, Ubuntu, Couchbase, Bash, AWS(Athena, Glue, Cloudwatch, S3, SQS) Google (Apps Scripts, Forms), Zapier, YouCanBookMe.WITH THE CLIENT (Online Privacy Company):• 460% time improvement (from 23 hours to 5) by designing and developing a data pipeline in AWS Glue and PySpark for 1 TB of data processing every day.• 50% time reduction (from 8 to 4 hours) on bulk data transfer activities by emulating AWS functionality (aws s3 cp command) with Spark and Scala solutions, moving 1TB of data daily.• Completed efficient and costless No-SQL solutions (in Couchbase) for sensitive data removal and fake data creation.AT DISTILLERY:• 25% time reduction in recruitment activities by orchestrating and implementing an end-to-end multidisciplinary cloud solution (Social Networks Recruitment) to contact and evaluate candidates through social media.• As a Data interviewer, I created and received amazing feedback regarding an end-to-end multidisciplinary cloud solution (Automated Quality control), for interviewer evaluation and takeaways.• Sped up both customer needs and the onboarding process for newcomers by constructing a nurtured “Data Science Career Path”. Show less

TECH STACK: Python (Pandas), SAS, Teradata, Spark (PySpark), MacOS, Bash, Jupyter, Google Apps Scripts.• 12% improvement on funded accounts by creating business rules based on analysis performed on customer likelihood of interest in commercial messages via email, SMS, and push notification.• 8% marketing campaign effectiveness (called Red Flags), launched for centennials, based on data published by the Mexican government regarding bank results. This was carried out by a… Show more TECH STACK: Python (Pandas), SAS, Teradata, Spark (PySpark), MacOS, Bash, Jupyter, Google Apps Scripts.• 12% improvement on funded accounts by creating business rules based on analysis performed on customer likelihood of interest in commercial messages via email, SMS, and push notification.• 8% marketing campaign effectiveness (called Red Flags), launched for centennials, based on data published by the Mexican government regarding bank results. This was carried out by a multidisciplinary team, having taken the orchestrator role. • Improved model explainability on Churn rate topics by suggesting synthetic variables based on Couple Therapy psychological results.• Recognized by BBVA Mexico’s Head of Client Solutions for the development of a market share prediction model (using Scikit-Learn and Genetic Algorithms). Show less

Mexico City Area, Mexico

TECH STACK: Bash, Scala, Ubuntu, RedHat, Jupyter, RStudio, Teradata, Windows 10, Spark (Pyspark), Cloudera (Cloudera Manager, Hue, Hive, Impala, Sqoop, Oozie).• 200% of Cloudera’s performance enhancement, measured by the number of users on the platform (from 30 to 60), by using time-and-resource-efficient solutions in PySpark, Bash, and R.• 100% data migration goals achieved (productive models and datasets, with their respective adjustments and validations) from Cloudera to… Show more TECH STACK: Bash, Scala, Ubuntu, RedHat, Jupyter, RStudio, Teradata, Windows 10, Spark (Pyspark), Cloudera (Cloudera Manager, Hue, Hive, Impala, Sqoop, Oozie).• 200% of Cloudera’s performance enhancement, measured by the number of users on the platform (from 30 to 60), by using time-and-resource-efficient solutions in PySpark, Bash, and R.• 100% data migration goals achieved (productive models and datasets, with their respective adjustments and validations) from Cloudera to Datio, having both Product Owner and Data Architect roles simultaneously.• Successfully emulated productive activities in a non-productive environment (Cloudera) by carrying out an analytical approach on technical topics (RAM and cores used, schedules) and user assessment programs regarding best software practices. Show less

Mexico City Area, Mexico

TECH STACK: Python (Pymongo, Flask), Ubuntu, MongoDB, Third-party tools. • Built webservices to perform ETLs and analysis of real-time data based on both company's and partners' APIs. Due to their Beta phase nature, I also tested and documented them.

Mexico City, Mexico

TECH STACK: Debian, Bash, Apache HTTP Server.• Created and maintained the infrastructure needed to properly display the project's website.

Mexico City Area, Mexico

TECH STACK: Python (Pymongo, Boto3, Pandas), Ubuntu, No-SQL (MongoBD), SQL (PostgreSQL, MySQL), AWS SQS.• 20% of speed increase on bulk data transfers (from MongoDB to both PostgreSQL and MySQL) with Bash scripts.• Received great comments regarding the re-design and implementation of an ETL architecture by using simple capabilities such as Linux utilities (cron, nohup), SQL/No-SQL scripts, Python (boto3), and AWS SQS.• Increased the understanding of stakeholders and… Show more TECH STACK: Python (Pymongo, Boto3, Pandas), Ubuntu, No-SQL (MongoBD), SQL (PostgreSQL, MySQL), AWS SQS.• 20% of speed increase on bulk data transfers (from MongoDB to both PostgreSQL and MySQL) with Bash scripts.• Received great comments regarding the re-design and implementation of an ETL architecture by using simple capabilities such as Linux utilities (cron, nohup), SQL/No-SQL scripts, Python (boto3), and AWS SQS.• Increased the understanding of stakeholders and technical and analytical teams by establishing a committee to assist with data-related technical issues and support their data infrastructure needs. Show less

Mexico City Area, Mexico

TECH STACK: R, C#, Python, Windows 8, SQL (Microsoft SQL Server), Microsoft Analysis Services.• 15% accuracy improvement (from 70% to 85%) on credit risk prediction models by using Genetic Algorithms for feature selection tasks.• 10% increase in speed on potential slow payer identification by using the “Rumor Mill” agent-based model.• Refactored a model generator system that compared performances among unaided-created models, enhancing model interpretability.

Mexico City Area, Mexico

TECH STACK: Netlogo.SUPERVISOR: PhD. Otto Hahn Herrera.• Supported students in programming techniques and modeling activities. • Responsible for developing students' assessments.

Mexico City Area, Mexico

TECH STACK: GAMS, Matlab, Octave, Python.SUPERVISOR: PhD. Adriana Espinosa Contreras.• Contributed to the development of a software product used to test the batteries of hybrid cars, which was part of Dr. Adriana's dissertation.• Translated and optimized simulation code from Matlab to Octave. • Included additional functionalities such as a graph module and a partial differential equation solver.

Mexico City Area, Mexico

TECH STACK: Microsoft Office (Access, Excel, Word).SUPERVISOR: B.S. Alejandrina Rivera de la Peña. • Responsible for S.A.I.C. 5.0, which is a statistical and economic software.• Updated databases VCA-1 to VCA-5 (for police reports), which included dashboard activities in Excel to ease the mentioned report visualization.• Provided equipment and software installation services.• Contributed with the Mayor’s speeches.

Master Of Science - Ms

Bachelor Of Science - Bs