me

I work as a Data Scientist at Siemens in Nuremberg, where I fine-tune foundation models for electrical load demand forecasting and build Python forecasting APIs that have grown stakeholder adoption by 20x — with dashboards delivering insights to 500+ users. My work spans the complete ML lifecycle: data engineering, model development, validation pipelines, and production deployment.

I am pursuing an MSc in Artificial Intelligence at Friedrich-Alexander-Universität Erlangen-Nürnberg (Note 1.6, expected 2026), building on a BE in Industrial and Manufacturing Engineering from NED University Karachi, where I graduated 2nd in class with a 3.9/4.0 GPA and received multiple merit-based scholarships from HEC Pakistan.

Across previous roles as a freelance analyst, and at Auvitronics, I have delivered deep learning pipelines, time-series forecasting models, SQL data pipelines, and Power BI/Tableau dashboards. My core stack: Python, PyTorch, TensorFlow, GenAI, REST APIs, Docker, and AWS.

Work Experience

DATA SCIENCE AND ANALYSIS

DEEP LEARNING

DATABASE

SPARQL microsoft-access
Version Control & Virtualization

Data Scientist

Siemens

Oct 2024 - present

Nuremberg, Germany

  • - Fine-tuned foundation models for electrical load demand forecasting, improving prediction accuracy and scalability across real-world energy datasets.
  • - Developed and optimized Python-based forecasting models, improving accuracy by 10–12% across industrial use cases.
  • - Engineered scalable forecasting APIs with validation pipelines, increasing adoption by 20x and enhancing model reliability.
  • - Delivered analytical dashboards to 500+ stakeholders, supporting planning, decision-making, and business alignment.

Data Scientist and Analyst

Freelance

October 2021 - March 2023

Remote

  • - Built time-series forecasting models for sales and inventory planning, improving demand planning efficiency.
  • - Developed SQL-based data pipelines and validated datasets to ensure high data quality and KPI reliability.
  • - Created Tableau and Power BI dashboards for automated reporting and performance monitoring across departments.

Data Analyst

Auvitronics Limited

October 2020 - October 2021

Karachi, Pakistan

  • - Collected, structured, and analyzed logistics and demand datasets, improving operational efficiency by 20%.
  • - Forecasted shipment trends and optimized transport routes, reducing costs by 10–15%.
  • - Prepared structured reports and communicated analytical findings to senior management for strategic decisions.

Education

FAU Erlangen-Nürnberg

Master of Science - MSc, Artificial intelligence

Grade: Grade 1.6

Courses:

  • Pattern Recognition and Analysis
  • Machine and Deep Learning
  • Explainable Machine Learning
  • Data Analysis and Visualization

NED University of Engineering

Bachelors - BE, Industrial and Manufacturing

Grade: Grade 3.907 (Equivalent German Grade 1.1)

Courses:

  • Manufacturing Processes
  • Machine Learning for Manufacturing
  • Data Analysis
  • Data Driven Processes Control

My Portfolio

img

Benchmark Creation for ML & Foundation Models on multi-domain Datasets

Consolidated benchmark creation for ML and foundation models across multi-domain datasets. Domain-specific finetuning and parameter-efficient finetuning (PEFT) for time-series foundation models including MOIRAI and Chronos2, enabling robust evaluation and adaptation at scale.

img

Domain-Specific Finetuning & PEFT for Time-Series Foundation Models

Parameter-efficient finetuning (LoRA/PEFT) of time-series foundation models including MOIRAI and Chronos2 for domain-specific electrical load forecasting. Delivers adapted models that outperform general-purpose baselines while training only a fraction of parameters.

img

Extrudely: 2D Engineering Drawings to 3D CAD Conversion

Building a vision-language model (VLM) workflow that automates the conversion of 2D engineering drawings into 3D CAD files. The system interprets orthographic projections, extracts geometric features and dimensions, and generates parametric 3D models — streamlining the CAD reconstruction pipeline for manufacturing and design workflows.

img

Background-Informed Object Labeling for Semi-Supervised Detection

Innovative background separation and aspect ratio analysis drive this project's semi-supervised object detection. Distinguish humans and vehicles by extracting data from video frames, pushing the efficiency boundaries of detection methods.
Animated GIF

xML Framework: Decoding Neural Networks on Pokémon Dataset

Explore neural network intricacies with our eXplainable Machine Learning (xML) framework on Pokémon data. Enriched with data augmentation, this project expands 800 images into 60,000, offering transparent AI insights through visualization and interpretation tools. Uncover model intricacies for a deeper understanding of your models.

Animated GIF

Real-time Facial Emotion Recognition Web Application

Our web app uses AI to analyze facial expressions in real time, helping users understand and categorize emotions instantly. Ideal for mental health, UX enhancement, and education.

img

Brain tumor diagnosis using SIFT-DNN

This project is for diagnosis of brain tumor with a dataset of four classes using SIFT and Deep learning architecture

img

YOLO v5 Object Detector Implementation

Successfully implemented a state-of-the-art YOLO v5-based object detection system, leveraging cutting-edge techniques for real-time and accurate identification of objects in images. This project showcases my expertise in computer vision and deep learning, ensuring optimal performance in object detection tasks.

img

Employee Recommender System

Unlock personalized career paths with the HR Recommender System, tailoring recommendations for employees based on their unique traits. Enhance workforce engagement and satisfaction through targeted professional development suggestions.

img

Employee Performance Analysis

Uncover valuable insights in employee performance through analysis of environment satisfaction, social satisfaction, and job satisfaction. Predict and enhance workforce productivity with data-driven precision.

img

Customer Churn and Risk Analysis

Analyzed customer churn and risk factors, delivering actionable insights through dynamic visualizations for informed decision-making.

img

Diversity & Inclusion in a Company

Implemented initiatives for fostering workplace diversity, leveraging data insights to enhance inclusion and strengthen organizational culture.

Get in Touch