Aimal Rehman

Data Scientist — Turing.com

Remote · Contract · Palo Alto, California, USA

• Developed high-quality Python and SQL training data for Google Gemini, covering end-to-end data analysis, machine learning, and production pipeline workflows.
• Created post-training data across SFT for teaching desired solutions, RM for learning human preferences, and RLHF for optimizing model behavior.
• Fine-tuned and evaluated multiple internal Gemini models, improving coding, analytical reasoning, instruction-following, and machine learning performance.

Senior Data Scientist & Tech Lead — Tkxel

Onsite · Full-time · Lahore, Pakistan

Executive Assistant GPT

• Led the development of an internal AI platform using HR, marketing, sales, and organizational data.
• Built a RAG-based company Q&A assistant with session memory, conversational context, and multi-step function calling.
• Delivered prompt-driven PowerPoint and case-study generators for creating client-ready sales and business-development content.
• Designed reusable data-ingestion and vector-search pipelines supporting multiple enterprise document formats.

Taxation Software Chatbot — Tax Back International

• Led the delivery of a Llama 3-based chatbot for a European tax-filing company, helping individuals and businesses navigate the platform and complete tax returns accurately.
• Built a multimodal ingestion pipeline for text and tabular data across PDF, DOCX, CSV, and Excel files, and optimized the supporting RAG workflow.
• Successfully completed the project, generating an additional $0.1 million in revenue for Tkxel.

Data Scientist — Alivia Analytics

Remote · Contract · USA

• Hired by SeeThruTec in Pakistan and outsourced to Alivia Analytics-USA.
• Led AI initiatives addressing healthcare fraud, waste, and abuse while supporting prepay and postpay payment integrity for multiple health insurance clients.
• Developed OCR-based pipelines to extract and structure information from medical records, clinical notes, and insurance claims.
• Built a document-comparison system that displayed similarity matrices and highlighted duplicated lines across provider records.
• Developed validation workflows to verify consistency between submitted claims and supporting medical documentation.
• Integrated an AI-based fraud detector to flag suspicious or potentially fabricated medical records for investigation.
• Helped identify approximately 3–5% of claims as high risk, preventing several hundred thousand dollars in improper payments for a major healthcare insurer.

Research Machine Learning Engineer — LUMS

Remote · Contract · Lahore, Pakistan

• Worked in the Clinical and Translational Imaging Lab, Department of Electrical Engineering, School of Science and Engineering, LUMS.
• Developed and evaluated U-Net, Conditional GAN, and Wasserstein Conditional GAN architectures for cardiac MRI synthesis and semantic segmentation.
• Built segmentation pipelines for the right ventricular blood pool, left ventricular blood pool, and myocardium across different phases of the cardiac cycle.
• Conducted rigorous model training, validation, loss-function design, and performance evaluation using PyTorch and TensorFlow.
• Investigated robust segmentation methods to improve anatomical-boundary detection and chamber and myocardial volume estimation.
• Enabled automated volume calculation during systolic and diastolic phases to support cardiac-function assessment.
• Designed the framework to assist radiologists with faster, more consistent diagnostic and prognostic evaluation.

Data Scientist — TechGenies LLC

Remote · Full-time · Islamabad, Pakistan

• Analyzed workforce data across multiple organizations to uncover insights into retention, performance, absenteeism, and staffing trends.
• Developed absenteeism forecasting pipelines using ARIMA for short-term prediction and Facebook Prophet for long-term workforce planning.
• Built employee churn models using Random Forest and Gradient Boosting, combining predictive accuracy with interpretable drivers of attrition.
• Applied OCR and NLP to develop a resume–job matching solution that automated candidate shortlisting and reduced manual screening effort.
• Designed HR analytics proofs of concept that secured clients, attracted project funding, and evolved into revenue-generating products and services.
• Supported recurring revenue through continued delivery and maintenance of resume matching, absenteeism forecasting, and employee churn solutions.

Data Scientist I — Afiniti

Remote · Full-time · Islamabad, Pakistan

• Analyzed client contact-center data and prepared data-integrity reports to identify issues affecting model performance.
• Trained and deployed Decision Tree and STAN-based models for real-time caller-to-agent pairing in production call centers.
• Monitored deployed models and evaluated production performance to identify opportunities for further improvement.
• Performed feature selection and engineering using Random Forest and LightGBM models to improve predictability.
• Grid-searched predictive pipelines to optimize performance on Sky UK's largest queue, contributing approximately £90,000 in monthly revenue gain.

Teaching Assistant — Information Technology University

Onsite · Contract · Lahore, Pakistan

Graduate course: Statistical and Mathematical Methods for Data Sciences

• Conducted weekly tutorials covering probability, statistics, linear algebra, and convex optimization for data science.
• Held office hours and guided students through mathematical concepts and problem-solving techniques.
• Prepared and evaluated assignments and quizzes, developed grading keys and course blogs, and maintained the course file.

Graduate Teaching Assistant — LUMS

Onsite · Contract · Lahore, Pakistan

Graduate course: Applied Probability

• Supported a graduate course offered by the Department of Computer Science, LUMS.
• Conducted weekly recitations on probability concepts applied to computer networks and data science.
• Held office hours and supported students with theoretical concepts and problem solving.
• Prepared and evaluated assignments and quizzes, developed grading keys, and maintained the course file.

Online Tutor — Mathematics & Physics

Remote · Freelance · United Kingdom

• Delivered online Mathematics and Physics lessons to O-Level, GCSE, and IGCSE students in the United Kingdom.
• Prepared and evaluated homework assignments, practice exercises, and examinations.

Research Assistant — LUMS

Onsite · Contract · Lahore, Pakistan

• Worked in the Electronics and Embedded Systems Lab, Department of Electrical Engineering, School of Science and Engineering, LUMS.
• Designed a hardware-friendly NOMA receiver for iterative multiuser detection and channel decoding.
• Applied Maximum Likelihood and MAP estimation with Monte Carlo sampling to optimize belief-propagation decoding over bipartite factor graphs.
• Developed channel-decoding methods for accurate transmitted signal recovery while minimizing transmission energy and computational complexity.
• Co-authored a peer-reviewed publication presented at the IEEE/ACM Design Automation Conference.