{"id":51604084,"url":"https://github.com/daniellevenstein/daniellevenstein","last_synced_at":"2026-07-12T00:02:08.350Z","repository":{"id":340620268,"uuid":"1166053789","full_name":"DanielLevenstein/DanielLevenstein","owner":"DanielLevenstein","description":"Data Engineer \u0026 AI Developer showcasing samples of work. ","archived":false,"fork":false,"pushed_at":"2026-06-21T17:21:30.000Z","size":4149,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-21T19:03:25.546Z","etag":null,"topics":["data-science","image-classification","machine-learning"],"latest_commit_sha":null,"homepage":"https://www.kaggle.com/code/daniellevenstein","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/DanielLevenstein.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-02-24T20:39:19.000Z","updated_at":"2026-06-21T17:21:33.000Z","dependencies_parsed_at":null,"dependency_job_id":"52280d44-7f06-492f-88d6-ec0b5fefbfce","html_url":"https://github.com/DanielLevenstein/DanielLevenstein","commit_stats":null,"previous_names":["daniellevenstein/daniellevensteinportfolio","daniellevenstein/daniellevenstein"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/DanielLevenstein/DanielLevenstein","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DanielLevenstein%2FDanielLevenstein","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DanielLevenstein%2FDanielLevenstein/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DanielLevenstein%2FDanielLevenstein/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DanielLevenstein%2FDanielLevenstein/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/DanielLevenstein","download_url":"https://codeload.github.com/DanielLevenstein/DanielLevenstein/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DanielLevenstein%2FDanielLevenstein/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":35377925,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-07-11T02:00:05.354Z","response_time":104,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["data-science","image-classification","machine-learning"],"created_at":"2026-07-12T00:02:03.597Z","updated_at":"2026-07-12T00:02:08.343Z","avatar_url":"https://github.com/DanielLevenstein.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"# Data Engineer \u0026 Cloud Developer\n\nSoftware engineer with 6+ years of experience working in AWS environments and data systems. Recently focused on machine learning and data analysis projects, including clustering, and business analytics.\n\n🔗 [Deployed Models](https://huggingface.co/DanielLevenstein)\n🔗 [Kaggle Notebooks](https://www.kaggle.com/daniellevenstein)\n\n## Professional Work:\n\n- Developed Java/Scala scripts to migrate and normalize legacy financial records into updated schema formats for downstream validation workflows. (FINRA)\n- Built a Lambda-based system to track application uptime and system health signals. (b.well)\n\n## Certifications:\n\n- Certified AWS Cloud Practitioner (AWS, 2025)\n- Post-Graduate Certificate in Artificial Intelligence \u0026 Machine Learning (UT, 2026)\n\n## Featured Projects:\n![Featured_Projects](charts/Featured_Projects.png)\n*Figure: Shows an image of my AWS Rag and AWS Certification coach project side by side.*\n\n## Detailed Information:\n🎓 AWS Certification Coach (ML Evaluation, Streamlit, Docker)\u003cbr\u003e\n🔗 GitHub: [DaielLevenstein/AWS-Certification-Coach](https://github.com/DanielLevenstein/AWS-Certification-Coach)\u003cbr\u003e\n🔗 Live Demo: [aws-certification-coach-onrender](https://aws-certification-coach-latest.onrender.com/)\n\n**Overview:** Developed an AI-powered study application for AWS certification practice. The app presents freeform AWS exam-style questions, evaluates learner answers with a trained local classifier, and returns structured coaching feedback with score, missing concepts, detailed answer guidance, and original multiple-choice provenance.\n\n**Key Features:**\n\n- Built a Streamlit study interface for answering AWS certification questions in freeform text.\n- Generated self-authored AWS exam-style training, holdout, and sample question datasets.\n- Designed a combined JSON artifact format that stores questions, original multiple-choice provenance, correct answers, wrong answers, and partial-credit examples together.\n- Trained a local answer classifier to evaluate full-answer correctness with release metrics for accuracy, precision, recall, and confusion matrix results.\n- Created a partial-credit regression model using continuous answer ratings and mean-squared error evaluation.\n- Added validation tests for classifier behavior, low-credit answer rejection, memory overhead, generated artifacts, and model performance.\n- Containerized the application with Docker for local deployment and Render-ready hosting.\n\n**Tools \u0026 Technologies:** Python, Streamlit, Docker, Machine Learning, Classification Models, Regression Models, JSON Data Pipelines, Automated Testing, AWS Certification Content\n\n---\n\n### 🎺 AWS Documentation RAG System (LLM, Vector Search, ETL Pipeline)\n\n🔗 [GitHub: Aws-Documentation-Rag](https://github.com/DanielLevenstein/aws-documentation-rag)\n\n**Overview:**\nDeveloped a retrieval-augmented generation (RAG) application for querying AWS documentation. Built a pipeline for crawling AWS documentation pages, preprocessing and storing data in a vector database, and generating technical responses using a language model through a Streamlit interface.\n\n**Key Features:**\n\n- Developed a web crawler for ingesting AWS documentation pages.\n- Kept track of feature downloads and crawl depth to prevent redownloading files repeatedly.\n- Implemented vector-based semantic retrieval for documentation search workflows.\n- Integrated a language model for context-aware technical question answering.\n- Built a Streamlit frontend for interactive querying and response generation.\n- Containerized the application using Docker for local deployment and testing.\n\nTools \u0026 Technologies: Python, Streamlit, Docker, Vector Databases, LLMs, Web Scraping, ETL Pipelines\n\n---\n\n### 🪖 Helmet Classification Pipeline (CNN, Training Data Curation)\n\n🔗 [Interactive Streamlit App](https://huggingface.co/spaces/DanielLevenstein/Helmet_CNN_Data_Quality_Case_Study)\u003cbr\u003e\n🔗 [Helmet CNN Training Notebooks](https://github.com/DanielLevenstein/Helmet_CNN_Model_Training_Notebooks)\n\n**Overview:**\nDeveloped an image classification pipeline for helmet detection using a CNN trained on standardized 100x100 image inputs. Build a data curation workflow to transform raw annotation images into standardized size and remove samples with too low a resolution to be useful.\n\n**Key Feature:**\n\n- Built four CNN-based image classification models using four different training datasets.\n- Evaluated models based on a holdout dataset consisted of equal numbers of samples from Dataset2, and Dataset3.\n- Improved accuracy of a classification model from 77% to 94%, and calculated precision, recal, and f1 metrics for each model.\n- Identified categories of images model struggles with for further fine-tuning.\n\nTools \u0026 Technologies: TensorFlow, Keras, OpenCV, Streamlit, HuggingFace\n\n---\n\n### 🍕 FoodHub – Delivery Data Analysis\n\n🔗 [GitHub: FoodHub Data Analysis](https://github.com/DanielLevenstein/FoodHub_UT_ML_Project1)\n\n**Overview:**\nDeveloped discount program for a food delivery app, which reaches up to 12% of the population under model conditions.\n\n**Key Features:**\n\n- Cleaned dataset and imputed missing ratings.\n- Identified top-rated restaurants by cuisine.\n- Calculated discount cost vs. total customer reach.\n\n![Discount Cost vs Population Reach](charts/Discount_Cost_vs_Total_Reach.png)\n\n*Figure: Discount cost vs total reach with value per customer analysis.*\n\n---\n\n### 💪 Stress vs. Physical Activity – Kaggle Data Analysis\n\n🔗 [Kaggle: Stress vs Physical Activity Correlation](https://www.kaggle.com/code/daniellevenstein/stress-vs-physical-activity-correlation)\n\n**Overview:**\nAnalyzed relationships between stress levels and physical activity using real-world survey data.\n\n**Key Features:**\n\n- Performed exploratory data analysis and correlation studies\n- Generated three population clusters using K-Means\n- Identified occupation-based stress correlations\n- Visualized cluster separation and activity distribution\n\n![Clustering Overview](charts/Stress_Level_by_Cluster_Side_by_Side.png)\n\n*Figure: K-means clustering reveals distinct population-level groups based on total activity and stress.*\n\n## 📫 Contact\n\nPhone: 512-885-5925\nLinkedIn: https://www.linkedin.com/in/daniel-aaron-levenstein/\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdaniellevenstein%2Fdaniellevenstein","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdaniellevenstein%2Fdaniellevenstein","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdaniellevenstein%2Fdaniellevenstein/lists"}