Harrinson Andres Arrubla Ramirez
Download
Recommend
Messages
Share
Harrinson Andres Arrubla Ramirez

Harrinson Andres Arrubla Ramirez  new_releases

About

Detail

Statistics and Data Science Master Student
Florida, United States
Contact Harrinson regarding: 
work
Full-time jobs
Starting at USD45K/year
Flexible work
id_card
Internships
Open to unpaid internships
connect_without_contact
Finding mentors
groups
Networking

Timeline

work
Job
school
Education
folder
Project (professional or personal)
auto_stories
Publication

Résumé

Jobs verified_user 0% verified
  • S
    Data Scientist
    Salaz Design
    Mar 2024 - Current (1 year 5 months)
    - Developed machine learning pipelines (using Python libraries such as TensorFlow, PyTorch, etc) for demand forecasting of construction services to predict project demand with 75% accuracy. This enabled the company to optimize resource planning, reducing project delays significantly (%40+). - Created and enhanced a unified SQL database by combining data from Google Analytics, Salesforce CRM, and AWS Redshift. This streamlined marketing, sales, and customer insights analysis, improving reporting efficiency by 40% and supporting more precise, data-backed decisions. - Built ETL pipelines to collect, clean, and standardize historical project data from various sources, loading it into centralized databases. This enabled the identification of t
  • O
    AI Trainer;
    Outlier
    Jan 2024 - Current (1 year 7 months)
    - Fine-tuned and optimized Large Language Models for a wide range of projects (LLMs, AI Chatbots, etc), delivering weekly performance improvements of up to 6% in prediction accuracy and 2% in inference time, culminating in a 35% overall efficiency gain. - Developed logical reasoning prompts that improved model inference accuracy by 23% across three reasoning methodologies. - Rigorously evaluate AI model outputs, identify areas for improvement, analyze performance metrics, provide actionable feedback, and explore novel prompting techniques to optimize model performance across domains.
  • S
    Data Analyst
    Salaz Design
    Dec 2023 - Mar 2024 (4 months)
    - Designed and implemented ETL pipelines to extract historical project data from multiple sources, transform it for consistency, and load it into centralized databases for analysis, enabling trend identification in pricing, project types, and customer segmentation. - Developed interactive Tableau dashboards by integrating transformed data from ETL processes, allowing stakeholders to track key project metrics (e.g., budget adherence, timelines) and make data-driven decisions during project review meetings. - Automated reporting workflows using Excel, Google Sheets, and SQL, reducing manual report generation time by 25% and ensuring timely delivery of project status updates.
  • American Statistical Association
    T
    Vice President
    American Statistical Association, The University of Texas Rio Grande Valley
    Mar 2022 - Aug 2022 (6 months)
    - Played a key role in developing and executing initiatives to promote statistical education and awareness among members and the UTRGV Student and Alumni community. - Worked collaboratively with other association members to plan and coordinate events, workshops, and meetings, fostering a vibrant statistical community within the university.
  • T
    Graduated Research Assistant
    The University of Texas at Rio Grande Valley
    Mar 2022 - Aug 2022 (6 months)
    - Tutored students from diverse U.S. institutions, enhancing collaboration through varying academic backgrounds. Contributed to the success of NSF-funded research by assisting in fostering a learning environment that supported the $324,072 grant. - Instructed students on deep learning libraries (TensorFlow, Keras, PyTorch), taking them from zero to hero in neural networks and machine learning on Matlab, R, and Python. - Reviewed and contributed to the writing of a peer-reviewed publication titled "Numerical simulations for fractional differential equations of higher order and a Wright-type transformation" in the Q1 journal Partial Differential Equations in Applied Mathematics (Vol. 11, 2024), a direct outcome of my tutoring in the NSF Res
  • University of Texas Rio Grande Valley
    Graduate Teaching Assitant
    University of Texas Rio Grande Valley
    Jan 2021 - Oct 2021 (10 months)
    - Tutor mathematics and statistics graduate courses for students along the University of Texas Rio Grande Valley faculties. - Tutored graduate-level Integral Transforms course (Spring 2021) and Linear Algebra (Summer 2021), grading assignments and managing Blackboard/Webwork platforms with 80% of tutoring attendees passing the course. - For pre-statistics courses at the high school level, I designed and organized course materials on WeBWorK, including custom problem sets, graphics, and topics aligned with course objectives.
  • University of Caldas
    Lecturer
    University of Caldas
    Jan 2020 - Dec 2020 (1 year)
    Developing curricula, scheduling activities, projects, and exams for computing engineering and health department for statistics, computer science, linear algebra, and calculus classes.
Education verified_user 0% verified
  • University of Texas Rio Grande Valley
    Master of Science (MS) in Applied Statistics and Data Science
    University of Texas Rio Grande Valley
    Jan 2021 - Dec 2023 (3 years)
    GPA 3.9/4.0
  • Colombia's National University
    Bachelor in Science (BS) of Mathematics
    Colombia's National University
    Jan 2014 - Apr 2019 (5 years 4 months)
    GPA 3.6/4.0
Projects (professional or personal) verified_user 17% verified
  • T
    Optimizing Predictive Accuracy for Fatal Heart Failure via Machine Learning
    The University of Texas Rio Grande Valley
    Jun 2022 - Aug 2022 (3 months)
    Applied penalized regression (lasso and elastic net method), linear regression, and SVM models to predict fatal heart failure using medical records. Tuned models using techniques like lasso regularization and mathematical kernels to optimize predictive accuracy. The lasso model achieved the highest accuracy, demonstrating the capability of sparse linear models for this task. The analysis identified key predictive features - serum creatinine, age, and ejection fraction.
  • T
    Fine-Tuning Transformer Networks for Sentiment Analysis with Natural Language Processing (NLP)
    The University of Texas Rio Grande Valley
    Feb 2022 - Apr 2022 (3 months)
    Code the best Natural Language Processing (NLP) algorithm to classify positive and negative reviews on real-life unstructured text datasets (Game Review). Findings: destilBERT transformer got 93% accuracy after tokenization, POS, and keyword extraction.
  • T
    Statistical Learning for Customer Response Modeling in Real-World Marketing Big Dataset
    The University of Texas Rio Grande Valley
    Jul 2021 - Aug 2021 (2 months)
    Developed models using logistic regression, random forest, and support vector machines to predict high-probability customers for future direct mail marketing campaigns in real-world big data. Trained the models on historical customer response data to identify prospective customers likely to engage positively with future campaigns.
  • T
    Regression and Classification of BIO data
    The University of Texas Rio Grande Valley
    Jun 2021 - Jul 2021 (2 months)
    Utilized SAS software to perform statistical analysis on a real-world dataset of resting blood pressure measurements collected longitudinally. Applied linear regression models to assess the relationship between blood pressure and other variables over time. Additionally, implemented logistic regression techniques to predict the likelihood of hypertension based on input variables. All research was made on SAS programming.
  • Colombia's National University
    Representation theory of Partially Ordered Sets with an Equivalence Relation. verified_user Verified experience
    Colombia's National University
    Jan 2018 - Apr 2019 (1 year 4 months)
    A. G Zavadskij, V. M. Bondarenko among others develop the Posets theory. This is a recent mathematic area in which there are several topics to research in. The principal research objective is the study of the representation of Partially Ordered Sets and the representation of partially ordered sets with an equivalence relation.
Publications verified_user 0% verified
  • T
    Probabilistic solutions of fractional differential and partial differential equations and their Monte Carlo simulations
    The University of Texas Rio Grande Valley
    Jan 2024
    The work in this paper is four-fold. Firstly, we introduce an alternative approach to solve fractional ordinary differential equations as an expected value of a random time process. Using the latter, we present an interesting numerical approach based on Monte Carlo integration to simulate solutions of fractional ordinary and partial differential equations. Thirdly, we show that this approach allows us to find the fundamental solutions for fractional partial differential equations (PDEs), in which the fractional derivative in time is in the Caputo sense and the fractional in space one is in the Riesz–Feller sense. Lastly, using Riccati equation, we study families of fractional PDEs with variable coefficients which allow explicit solutions. T