Senior Machine Learning Engineer (Tabular… at PointofIT · Islamabad

Company: PointofIT
Location: F-5, Islamabad, Pakistan (On-site)
Experience Level: 4+ YearsAbout the RolePointofIT is seeking a hands-on, mathematically rigorous Senior Machine Learning Engineer who specializes natively in structured tabular data training, classification, and time-series forecasting.Note: This is NOT a Generative AI, RAG, or Prompt Engineering role. We are looking for an expert who lives and breathes rows, columns, feature engineering, and traditional statistical machine learning algorithms to drive core predictive business outcomes.Key Responsibilities

• Predictive Modeling: Design, train, and deploy production-grade machine learning models for tabular data challenges (e.g., risk scoring, fraud detection, demand forecasting, customer churn).
• Advanced Feature Engineering: Perform deep exploratory data analysis (EDA), handle complex missing value imputations, address extreme class imbalances, and engineer highly predictive features from massive datasets.
• Pipeline Development: Build and maintain scalable, robust data preprocessing and model training pipelines using SQL and Python.
• Model Evaluation & MLOps: Evaluate model performance using strict statistical frameworks (F1-Score, ROC-AUC, MAE, RMSE). Monitor and manage model drift and optimization in production.

Required Technical Toolkit

• Core Algorithms: Deep, hands-on expertise with gradient-boosted trees and ensemble methods (XGBoost, LightGBM, CatBoost, Random Forests), alongside classic regression and classification frameworks.
• Data Libraries: Expert proficiency in Python, Pandas, NumPy, and Scikit-Learn.
• Databases: Advanced SQL capabilities (PostgreSQL, MySQL, or SQL Alchemy) to structure, clean, and manipulate large relational databases.
• Infrastructure & MLOps: Solid experience with Docker, version control (Git), and model tracking frameworks (e.g., MLflow, DVC) to deploy models into production.
• Big Data (Plus): Experience with PySpark or Elasticsearch for large-scale data manipulation is a major plus.

What We Are NOT Looking For

• Candidates whose recent 2 years of experience are solely focused on building chatbots, fine-tuning LLMs (GPT, Llama), setting up RAG systems, or orchestrating LangChain/LangGraph workflows.

What We Offer

• Competitive salary package based on depth of mathematical and engineering expertise.
• Opportunity to lead high-impact predictive analytics engines for international enterprise systems.
• A collaborative environment at our prime office space in the Evacuee Trust Complex, F-5, Islamabad.

Pay: From Rs250,000.00 per month

Application Question(s):

• Which specific Python libraries and machine learning algorithms do you have the most hands-on experience using for tabular classification or regression?
• When working with a highly imbalanced tabular dataset (e.g., 95% majority class, 5% minority class), what exact techniques or algorithm-level parameters do you use to balance the model?
• Briefly describe a production-level model you trained where you had to perform extensive feature engineering on rows and columns. What was the business problem and the resulting model metric?
• This position is 100% pure on-site at our office in F-5, Islamabad, with working hours from 12:00 PM to 9:00 PM. Are you fully comfortable with this location and shift timing?
• What is your expected monthly salary in PKR for this role?
• What is your joining time / notice period if you are selected for this position?
• This role focuses strictly on Tabular Data (SQL tables, Excel, transaction forecasting). If your primary expertise is in Computer Vision (YOLO, OpenCV, Image/Video) or GenAI/LLMs, please answer No. Do you have hands-on experience training tabular models (XGBoost, LightGBM, Scikit-learn)?

Work Location: In person