Kevin is an experienced ML Engineer specializing in end-to-end machine learning pipeline implementations and AI solutions. With a solid foundation in Python and C++, he has successfully developed applications in healthcare, including melanoma and hair loss detection systems. His expertise extends to generative AI technologies, where he has integrated solutions using OpenAI and LangChain. Kevin is committed to delivering high-quality results and has consistently achieved measurable performance improvements in his projects.
Increased Customer Retention by 12.7% through a predictive model with an F1-Score of 85%.
Achieved 95% Accuracy in generating and visualizing pink Nitrogen Plasma using OpenCV techniques.
Reduced Network configuration latency by 50%, enabling faster and more responsive real-time network setups.
Maintained system stability with zero critical failures reported over a six-month period.
Overview: Developed a SAAS platform for doctors to provide prescriptions with speech-to-text analysis. Responsibilities: Utilized the Deepgram API to convert speech into text and tailored a model for healthcare-specific response generation. Implemented FastAPI and socket programming for frontend integration.
Key outcomes:
Developed a SAAS-based platform for doctors to provide verbal prescriptions and details.
Overview: Developed a system leveraging deep learning and image processing to diagnose hair conditions from varied angles. Responsibilities: Utilized CNN and ResNet 152 models, achieving 97% accuracy for no hair loss cases. Implemented privacy measures by blurring facial features for user anonymity before storing images in S3.
Key outcomes:
Achieved 97% accuracy for no hair loss cases using CNN and ResNet 152 models.
Overview: Implemented an end-to-end Machine Learning pipeline for melanoma cancer detection, using generated images. Responsibilities: Constructed CNN models with Data Augmentation and hyper-parameter tuning to reduce overfitting.
Key outcomes:
Achieved 81% Accuracy and 87% Recall for melanoma image classification.
Key outcomes:
Analyzed YouTube comments using sentiment analysis and OpenAI for content summarization.
Identified negative/positive comments and assessed humor/tone scores.
Key outcomes:
Developed an application for personalized product recommendations based on user input and images.
KEVIN
ML Engineer