Victor David Capriles Hinojosa Email and Phone Number

Creating software solutions that incorporate AI algorithms to solve specific problems.

- Developed POC of a Retrieval-Augmented Generation (RAG) model to enhance the accuracy and reliability of the LLM Meta-Llama-3.1-8B model for precise retrieval and generation of documentation, tailored specifically for the QA department's needs. Deployed using chainlit's framework and added features such as, user session, chat persistance and memory context.- Streamlined the monthly KPI reporting process by leveraging a JIRA add-on integration within Google Sheets and Excel. Automated data extraction and optimized inter-spreadsheet communication through advanced formula scripting, resulting in a 50% reduction in report generation time.- Developed a master delivery sheet integrated with individual project owner sheets using IMPORTRANGE and QUERY functions. This setup centralized data collection, consolidating shipment units, delivery dates, and complaints into a unified tracking system and reduced data collection time by 60%.- Supervised software release processes across multiple projects in the healthcare and automotive industries, ensuring compliance with ISO 13485, Automotive SPICE, ISO 9001 or IATF 16949 standards. Managed end-to-end documentation, traceability, and process adherence from development through to production readiness.

- Developed POC of an Image Classifier able to distinguish the part quality defect "Excess Energy" in 3d printed parts for the MultiJet Fusion 5200, using Deep Learning to train in Edge Impulse the model and the transfer learning tool to use the pre-trained CNN MobileNetV2.- Aquired the data, labeled the images in the datasets and allocated the images with a ratio of 80% : 20% for training and validation. Performed decision making regarding fine tuning of hyper-paramaters (Epochs, Learning Rate) and Confusion Matrix evaluation.- Deployed the trained model into a Raspberry Pi and performed inference with an accuracy of 88.8%.- Generated a monthly report for stakeholders by leveraging SQL to extract and analyze key metrics on printer usage, powder consumption, connectivity, and printer fleet number, providing data-driven insights of installed printers from the field.- Learned the different partquality defects that can appear in 3d printed parts, their root cause and possible solutions for the MultiJet Fusion 4200 & 5200. Worked with differents materials and their processing and post processing workflows, such as PA 12, PA 12GB, PA 11, TPU and TPA.

-Taught students from ages 7-12 by following a structured guide which helps in the correct luinguistic development of the students.- Wrote quarterly reports with detailed information about the evolution of the student for the parents.- Interacted with the parents in meetings to review the student's performance through the couse.- Developed classroom management skills and social skills to be able to communicate and interact with colleagues, parents and most importantly, students.