Instacart Customer Behavior Analysis
Deep EDA on grocery ordering patterns, repeat purchases, order timing, product affinities and customer segmentation.
- Python
- Pandas
- Matplotlib
- +1
AI Engineer, Data Scientist, AI Educator
I'm Ahmad, a PharmD graduate who taught himself machine learning and now builds production AI and MLOps systems. I've mentored 100+ students from medicine, business, engineering and the arts into data and AI roles.
or message me on WhatsAppAI with AhmadMy free learning communitySix stage-by-stage roadmaps, from your first line of Python to deployed AI products, with lessons, notebooks, portfolio projects and progress tracking. Free for everyone.learn.ahmadhassanrx.com
5+
years in data science
1,500+
live teaching hours
100+
students mentored
30+
end-to-end ML projects
About me
I trained as a pharmacist at Quaid-i-Azam University, and midway through my PharmD I discovered data science. I taught myself machine learning in the evenings, then took on freelance data science work on Fiverr and Upwork.
After graduating with a 3.9 cGPA I moved fully into AI: consulting on ML projects, tutoring data science on Superprof, and interning in AIOps at Al Nafi, where I completed an RQF Level 6 Diploma in AIOps with an A*.
My route into AI was the one any CS student takes: programming, computer systems and the mathematical foundations of machine learning, so I teach the fundamentals properly, not just the tools. Today I build production AI and MLOps systems, and I've mentored 100+ students, most of them from medicine, business, engineering and the arts, into data and AI roles.
Next, I'm heading for a Master's and PhD in computational biology and bioinformatics, with one goal: AI-assisted drug discovery and development. It brings my training in medicine together with AI engineering and operations, and it is where my specialisation and my future lie.
Current focus
From a pharmacy degree to production AI, one deliberate step at a time, and on towards AI for drug discovery.
2020
Started teaching myself ML midway through the PharmD, then worked through the same foundations a CS student does: programming, computer systems and the maths behind AI.
2022
Client projects on Fiverr and Upwork.
2024
Graduated from Quaid-i-Azam University.
2024
Independent consulting, data science tutor on Superprof and AIOps intern at Al Nafi.
2026
Cloud, DevOps, security and AI systems.
A Master's and PhD towards AI-assisted drug discovery and development, bringing medicine and AI together.
PharmD
Quaid-i-Azam University
RQF Level 6 Diploma in AIOps (EduQual)
Al Nafi International College
Kubestronaut, AWS & Azure certifications
Cloud and Kubernetes certifications for AI engineering
Master's & PhD: computational biology and bioinformatics
Towards AI-assisted drug discovery and development
Skills
The stack I build with and teach, from the first notebook to monitored systems in production.
Flagship product
Translating clinical domain knowledge into a full-stack AI system: the kind of depth that comes from understanding both the medicine and the machine.
My flagship product: clinical decision support that helps physicians make the right prescribing choice by weighing the complete patient profile, with every recommendation backed by documented proof from clinical guidelines and rules.
Details shared selectively. Happy to discuss it in interviews and collaborations.
Tech stack
Learning roadmaps
AI Engineering is the flagship; the other five feed into it. Every roadmap lives on AI with Ahmad, free, with its lessons, notebooks, projects and progress tracking.
AI with Ahmad
Six stage-by-stage roadmaps, from your first line of Python to deployed AI products, with lessons, notebooks, portfolio projects and progress tracking. Free for everyone.
Portfolio
Over 30 projects spanning data analysis, machine learning, generative AI and BI, each built to solve a real problem. Healthcare is where I started, so it is where much of my work lives.
Deep EDA on grocery ordering patterns, repeat purchases, order timing, product affinities and customer segmentation.
Statistical analysis of two prepaid plans (Surf vs. Ultimate) to determine which generates more revenue for the business.
Multi-dimensional analysis of SaaS product usage, customer segments and retention patterns across Microsoft product lines.
Geographic and temporal analysis of rental markets in SF, uncovering pricing trends, neighborhood patterns and investment opportunities.
Analysis of ride patterns, peak hours, driver efficiency and revenue optimization strategies for a taxi service.
Sales trend analysis across gaming platforms, genres and regions to identify the most profitable market segments.
Clinical data analysis identifying key risk factors for stroke, including hypertension, BMI, glucose levels and age correlations.
Ensemble model (XGBoost + Random Forest) to predict which customers are likely to leave Beta Bank, achieving 85%+ F1-score.
Multi-stage regression model predicting gold recovery rate at rougher and final purification stages to optimize extraction.
Geospatial ML model using bootstrapping to identify the most profitable oil well locations, balancing risk and expected return.
Gradient boosting model for a used car marketplace to predict fair market prices, compared across LightGBM, XGBoost and Random Forest.
Time series forecasting of hourly taxi demand at airports using ARIMA, LightGBM and feature engineering on lag and window features.
Linear regression and obfuscation-safe model for predicting insurance payouts while protecting customer data privacy.
Decision tree and random forest classifier to recommend the optimal mobile plan (Smart/Ultra) based on user behavior.
End-to-end ML pipeline predicting the probability of loan default, with model explainability using SHAP values.
ML model predicting equipment failure probability to shift from reactive to proactive maintenance, reducing operational costs.
Predicts stroke risk from patient health records, handling heavy class imbalance with resampling and boosted trees.
A multiclass lung cancer risk model served as a Dockerised Flask web app that generates a PDF report.
Remaining Useful Life (RUL) prediction for jet engines using the NASA CMAPSS dataset, deployed as a Flask safety application that alerts maintenance teams before critical failure.
Deep learning approach to customer churn: an ANN with dropout regularization outperforming classical ML baselines.
A deep learning image classifier that grades MRI brain scans across four stages of dementia.
CNN-based image classification pipeline with transfer learning (ResNet), data augmentation and production-ready inference.
NLP pipeline using transformer-based models to automatically summarize customer complaints, reducing support team workload.
An LLM assistant that explains a patient's prescribed medicines, nutrition and lifestyle plan against international clinical guidelines.
Fine-tuned DistilBERT for multi-class complaint categorization, achieving >92% accuracy vs. a zero-shot baseline.
A nutrition-disorder specialist chatbot that answers from a medical reference library with retrieval-augmented generation.
DreamBooth fine-tuning pipeline to generate personalized images with Stable Diffusion, including custom artist style training.
Complex SQL queries on the AdventureWorks database: joins, window functions, CTEs and performance optimization.
Full ETL pipeline extracting from SQL Server, transforming with Python and Pandas, and loading to Power BI for executive dashboards.
Interactive Tableau dashboard visualizing employee attrition patterns, satisfaction scores and retention risk factors.
Tableau analysis comparing performance across medical settings: patient outcomes, treatment efficacy and resource utilization.
No projects in this category yet.
Open source
Fine-tuned models, experiment checkpoints and research artifacts published openly, from NLP and generative AI projects to domain-specific adaptations in the clinical and analytical space.
View my Hugging Face profileExplore models & spaces
Milestones
A track record built through years of consistent learning, teaching and building.
3.9 / 4.0 cGPA · Graduated 2024
Top-performing graduate in one of Pakistan's most rigorous pharmaceutical programs, building a foundation in scientific research, data interpretation and analytical thinking.
Al Nafi International College · Grade A*
An AIOps diploma covering cyber security, DevOps, cloud architecture, system administration and AI systems: the complete operational stack for modern AI deployment.
Superprof · Top tutor
Over 1,500 hours of live one-to-one and group instruction in data science, ML and AI, making complex topics accessible to complete beginners.
Cross-domain · Career transitions
Worked with over 100 students from medicine, business, engineering and liberal arts backgrounds, guiding them through career transitions into data science and MLOps roles.
KCNA · CKAD · CKA · CKS · KCSA
Pursuing the full Kubestronaut certification stack, from cloud native associate through security specialist, as part of the Golden Kubestronaut program.
CISSP · CCSP · CCSK · ISSAP
Studying toward top-tier security certifications covering enterprise security architecture, cloud security and identity management at the advanced practitioner level.
Production-grade · Multi-domain
Delivered 30+ complete ML projects spanning data analysis, machine learning, deep learning, NLP, computer vision, GenAI and business intelligence.
Al Nafi International College · Presented to Mohammad Faisal
Live-streamed final year project defense for the EduQual RQF Level 6 AIOps programme: "Orchestrating a Precision Agriculture Platform with Drone Analytics, IoT Sensors, and Machine Learning for Sustainable Farming", delivered before an examination panel.
Self-taught → industry practitioner
Built expertise from zero in 2020 through deliberate practice, real projects and hands-on teaching.
Get in touch
Whether you're looking to hire an AI engineer, collaborate on an AI project or research, or want guidance on your own path, I'd love to hear from you.