Tashfeen Ahmed

Data Analyst

Focused on analytics, BI, machine learning, data engineering, and GenAI

McGill MMA graduate with 3+ years of experience in healthcare analytics, reporting, pipeline automation, and applied machine learning.

View featured work All projects Send message
Analytics and BI

Power BI dashboards, KPI design, A/B testing, and decision support

Applied Machine Learning

Supervised and unsupervised learning, forecasting, causal inference, and NLP models

Data Engineering and GenAI

SQL, PySpark, Microsoft Fabric, Databricks, ETL automation, and LLM-powered workflows

Project Highlights

Selected projects

GenAI

Natural Language to SQL

Built a LangChain and Ollama workflow that converts natural-language questions into validated SQL queries for safer self-service analytics.

LangChain Ollama SQL
Open project

BI

Dashboard Analytics

Built dashboard analytics projects focused on KPI tracking, business storytelling, and decision-ready reporting.

Power BI Dashboards Reporting
Open project

Causal Inference

Causal Inference for Bank Subscriptions

Estimated treatment effects from marketing engagement variables using uplift models, S-learners, T-learners, and SHAP-based interpretation.

Causal Inference XGBoost SHAP
Open project

AI Agent

Portfolio AI Agent

Built a RAG-based AI agent for this website that evaluates project fit, retrieves the closest portfolio context, and returns short privacy-safe answers with relevant GitHub work.

Cloudflare RAG AI Agent
Open project

GenAI

Clinical Note Information Extraction

Structured clinical note content into extractable fields with an LLM-based workflow, making unstructured text easier to review and analyze.

Clinical NLP LLM Extraction JSONL
Open project

GenAI

Healthcare Chatbot

Fine-tuned Phi-3 with LoRA for supportive mental health dialogue, demonstrating efficient domain adaptation without full-model retraining.

Phi-3 LoRA Fine-Tuning
Open project

NLP

Predicting Job Salary from Job Description

Built an NLP classification project that predicts salary bands from job descriptions using TF-IDF features and a Naive Bayes model.

NLP TF-IDF Naive Bayes
Open project

Unsupervised Learning

Kickstarter Success Classification and Clustering

Combined classification and clustering to segment Kickstarter campaigns with K-means and DBSCAN and surface patterns in campaign goals, duration, and fundraising behavior.

Unsupervised Learning K-Means DBSCAN
Open project

Forecasting

Demand Forecasting with Time Series

Compared ARIMA and Prophet on historical sales demand, evaluated forecast quality, and used time-series decomposition to explain demand patterns.

ARIMA Prophet Time Series
Open project

Supervised Learning

PyTorch Multi-task Learning for House Prices and Categories

Built a multi-task deep learning model that handled price prediction and category classification in a shared supervised learning pipeline.

PyTorch Regression Classification
Open project

Technical Toolkit

Core tools and methods

Programming and Querying

Python SQL R PySpark Git

Analytics and BI

Power BI A/B Testing DAX Power Query Embedded BI MicroStrategy

Machine Learning

Supervised Learning Unsupervised Learning Causal Inference Time Series XGBoost LightGBM CatBoost Random Forest PyTorch

Data Engineering and Cloud

ETL/ELT Fabric Databricks Azure GCP Docker MLflow FastAPI Streamlit

GenAI and NLP

GenAI NLP LangChain Ollama Hugging Face LLM Fine-Tuning Embeddings TF-IDF

Projects

More GitHub projects

Loading projects from GitHub...

Contact

Get in touch