Niche Aim Technologies

•	Applied Visual Studio Code and Python 3.10 with libraries such as Pandas, Re, and Google cloud clients to monitor and detect potential anomalies in Google Cloud’s Big Query and Fire Store databases using the created data quality rules. The developed methodology compared information corresponding to the same object but stored in different databases by applying a set of conditions to determine entries with non-conformances. Once all the information was compared between the databases, all errors or anomalies were exported as CSV to a Big Query table and used in a Tableau dashboard to visually present the findings.

JR Data Engineer

University of Puerto Rico-Mayaguez

•	Performed as a teacher assistant, responsible for the undergraduate course ININ 4078: “Statistical Quality Control” laboratory sessions. Prepared laboratory material and assignments for topics such as: Quality Control Charts, Capacity Analysis, Exponentially Weighted Moving Average, Hotelling T2 Control Charts and Gauge R&R. Incorporated the use technical tools, such as Minitab and Excel, to facilitate quality analysis. Actively maintained grading up to date and provided office hours for student’s assistance.

Statistical Quality Control Teacher Assistant

Master Data Analysis

•	Utilized NumPy, Pandas, and Scikit-learn in Python 3.7 within Azure ML to assist a client with their custom feature engineering and feature selection. Evaluated correlated features from the custom process’s created new features to remove redundant variables. Applied data transformations and custom stepwise feature selection. With Azure ML, different classification models were compared, and the best performer was deployed via an API. The metric of interest was the recall score, which was on average above the target score of 70%.
•	Used Python 3.7 with Async, Pandas, NumPy, Openpyxl, and Win32com to assist a client with their supply chain department’s needs for automation of their reports. The script consolidated data from SAP, asynchronously merged all tables into one, and applied internal rules to perform feature creation. The created script completed the task of synthesizing the data, validating merges, and assigning the rules for feature creation 95.83% faster than the previous manual process.

Analytics Consultant

Honeywell

•	Used R Studio with readxl, data.table, zoo, SnowballC, tm, ggplot2, gbm, caret, e1071, randomForest, and inTrees to explore an initial supervised learning pipeline with parameter tuning. As part of the pipeline, the Random Forest, Support Vector Machine, and Gradient Boosted Machine algorithms were tuned, evaluated, and compared in multi-class classification tasks. Furthermore, different intermediate statistical methodologies for text mining were explored. The final tuned model’s metrics were then visualized and reported to the team. The outcome resulted in a pipeline with a mean accuracy above 80%, achieving the desired requirement.

Data Science Intern

•	Utilized SAP Hana as a data source and VBA macros that retrieved data given end-user parameters. Once the data was fetched, the VBA macro created Power Queries that refreshed pivot tables and estimated the bill of material forecast.
•	Developed reports and visualization using Power BI along with R for the creation of custom visuals and Node.js to package the visuals with the main goal of displaying relevant information about non-conforming materials and material tracking, such as new metrics, and historical data with filtering capabilities.

Supply Chain - Smart Manufacturing

Cell Manufacturing Technologies (CMaT)

•	Continuation research on "Exploratory Data Analysis for Transcriptomics" focused on Chimeric Antigen Receptor T (CAR-T) Cell Experiments, funded by the NSF Engineering Research Center for Cell Manufacturing Technologies (CMaT), with the intention of identifying differentially expressed genes associated with four different CAR-T constructs across CD4 and CD8 T cells, using the data and article published by Zhao Z et al. in “Structural design of engineered costimulation determines tumor rejection kinetics and persistence of CAR T cells” to determine molecular patterns that characterize the construct with highest T-cell persistence.
•	Employed the previous pipeline for “Exploratory Data Analysis for Transcriptomics” with the addition of the Random Forest supervised learning algorithm to use its Mean Gini Variable Importance to select the top 10 numbers of genes using the randomForest library to reduce the number of differentially expressed genes for data interpretation.

Undergraduate Student Researcher

•	The undergraduate research “Exploratory Data Analysis for Transcriptomics”, was funded by the NSF Engineering Research Center for Cell Manufacturing Technologies (CMaT), executed using R Studio, leveraging data sourced from the Gene Expression Omnibus (GEO) using GEOquery with the objective of investigating factors and conditions found in public transcriptomic experimental data and exploring differentially expressed genes across multiple datasets.
•	Performed data transformations for further analysis and creation of visuals using dplyr, tidyr, and data.table. Executed feature selection using Empirical Bayes to filter out genes with a p-value adjustment above the fixed thresholding set of 1% using the limma library. Integrated Gene Ontology (GO) and KEGG pathway analysis on the identified genes and leveraged state-of-the-art visualization techniques to offer a representation of results using clusterProfiler, DOSE, KEGGREST, pathview, enrichplot, and KEGGgraph.

Retreat Conference Presentation

Holding a bachelor's degree in industrial engineering from the University of Puerto Rico Mayaguez, accompanied by Microsoft certificates in artificial intelligence and data science. Passionate about applications, visuals, and methodologies in operations research, industrial engineering, data science, analytics, machine learning, and artificial intelligence. Experienced in supporting engineering and business teams with data-related topics including data/process automation, data visualization, and analytics for decision-making.

Taplio

Carlos Adames's Linkedin Analytics

Carlos Adames

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Carlos Adames's Linkedin Analytics

Carlos Adames

Want to drive more opportunities from LinkedIn?

Carlos Adames's Linkedin Analytics