Data & AI Portfolio
Hi, I’m Yenlik — I build practical data and AI projects.
I turn messy real-world information into clear decisions using Python, SQL, machine learning, NLP, forecasting, dashboards, testing, and responsible AI.
My work focuses on healthcare analytics, trust and safety, AI operations, forecasting, sustainability data, and operational intelligence.
Python | SQL | Machine Learning | NLP | Forecasting | AI Operations | Trust & Safety
Featured Projects
These projects demonstrate practical machine learning, NLP, analytics, deployment, testing, responsible AI, and business-focused problem solving.
Patient Feedback Intelligence App
A deployed NLP app that classifies NHS-style patient feedback into operational themes, helping teams identify issues in communication, waiting times, staff attitude, environment, and care quality. Built with Python, TF-IDF, Logistic Regression, Streamlit, pytest, MLflow, CI/CD, and responsible AI documentation.
Trust & Safety Risk Triage
A synthetic AI operations project that analyses user reports, identifies risk signals, prioritises cases, and supports operational decision-making for trust and safety teams. Built to demonstrate skills relevant to AI operations, platform safety, policy enforcement, analyst workflows, and risk-based triage.
Construction Safety NLP
An NLP project analysing construction safety narratives to identify recurring themes, risk patterns, and operational insights from unstructured text. Built with Python, text preprocessing, topic modelling, BERTopic, visual analysis, and business-focused recommendations.
More Projects
NHS Referral Demand Forecasting
Forecasting NHS Trauma & Orthopaedics waiting-list demand over a 6-month horizon using ARIMA and LSTM. Evaluated on a 6-month hold-out (Oct 2023 – Mar 2024). Supports NHS elective recovery capacity planning.
Student Retention Risk Modelling
XGBoost classification model identifying high-risk student cohorts in UK higher education using HESA-structured data. SHAP explainability surfaces top drivers for student support intervention.
Building Energy Anomaly Detection
Applied IQR, Isolation Forest, and One-Class SVM to ~1 million UK EPC records to identify likely retrofit priorities, with ~78% method agreement.
Health-Focused Retrofit Prioritisation
Identifies English local authorities where housing energy inefficiency, fuel poverty and respiratory health risk overlap most strongly. Composite priority score across 296 LADs.
Stack & Skills
Languages
Python, SQL, HTML, CSS
Data Science & Machine Learning
Pandas, NumPy, scikit-learn, Logistic Regression, ARIMA, SHAP, BERTopic, TF-IDF
NLP & AI
Text classification, topic modelling, risk triage, responsible AI, model cards, evaluation
Analytics & Dashboards
Power BI, Excel, Matplotlib, Plotly, KPI reporting, dashboard design, operational insights
Engineering & Deployment
GitHub, Streamlit, Hugging Face Spaces, pytest, CI/CD, MLflow, Makefiles
Domains
Healthcare analytics, trust and safety, AI operations, sustainability, forecasting, research operations
About Me
I am a data-focused professional with experience in research operations, health and safety coordination, procurement, project support, and business-facing problem solving at the University of Nottingham.
I build practical analytics and AI projects that turn messy real-world information into clearer decisions across healthcare, trust and safety, sustainability, forecasting, and operational intelligence. My work combines Python, SQL, machine learning, NLP, forecasting, dashboards, testing, deployment, and responsible AI documentation.
I am especially interested in roles where data, AI, operations, and real-world decision-making meet — including AI operations, trust and safety, healthcare analytics, responsible AI, and product analytics.
Roles I’m Targeting
Let’s Connect
I’m open to roles in AI Operations, Trust & Safety, and Responsible AI; Data Analysis and Product & Operations Analytics; and Healthcare & Sustainability Analytics.