Experience: 5–8 years of experience, with at least 3 years focused on machine learning engineering or applied science
Minimum Education: Bachelor’s degree in Computer Science, Mathematics, Statistics, or a relevant STEM field. Master’s or PhD degrees are highly preferred.
Responsibilities
Model Research, Development & Experimentation
- Design machine learning experiments systematically — from problem formulation and hypothesis development to algorithm selection and rigorous evaluation.
- Build and train large-scale ML models across supervised, unsupervised, and reinforcement learning paradigms according to business use cases.
- Perform LLM fine-tuning and customization using techniques such as LoRA, QLoRA, and PEFT for domain-specific applications.
- Design and implement accurate and reliable RAG (Retrieval-Augmented Generation) systems for knowledge-intensive applications.
- Conduct comprehensive model evaluation and benchmarking across accuracy, fairness, robustness against adversarial inputs, and production performance.
Data Engineering & Feature Development
- Design and implement scalable data preprocessing and feature engineering pipelines.
- Collaborate with Data Engineers to ensure data availability, quality, and consistency for training purposes.
- Develop data labeling strategies and quality assurance frameworks for training datasets.
- Identify and address data-related issues including dataset bias, imbalanced distributions, and data leakage.
Model Optimization & Production Readiness
- Optimize models for production environments by reducing model size (compression, pruning, quantization) while maintaining performance quality.
- Collaborate with MLOps Engineers to ensure models can be efficiently deployed, monitored, and retrained.
- Conduct profiling and optimization of inference performance, including latency, throughput, and memory footprint.
- Build automated evaluation pipelines to support continuous model quality assessment.
AI Research & Innovation
- Stay updated with and implement state-of-the-art techniques from recent research papers relevant to business needs.
- Evaluate and determine when to fine-tune proprietary models versus leveraging existing foundation models.
- Build internal knowledge bases documenting ML architecture decisions, experiment outcomes, and lessons learned.
- Collaborate with AI Engineers and AI Product Managers to translate model capabilities into tangible product solutions.
AI Ethics & Responsible ML
- Conduct bias audits and fairness assessments before model deployment into production environments.
- Implement model explainability approaches (SHAP, LIME, or other interpretability techniques) for use cases requiring transparency.
- Create comprehensive model cards documenting capabilities, limitations, intended usage, and known constraints.
- Ensure models comply with the organization’s responsible AI principles and governance standards.
Qualifications & Requirements
Must Have
- Strong mathematical foundation including linear algebra, calculus, probability & statistics, and optimization theory as core ML fundamentals.
- Advanced proficiency in Python with hands-on experience using PyTorch or TensorFlow in real production environments.
- Hands-on experience building and deploying ML models across at least 2–3 different use cases with measurable business impact.
- Deep understanding of machine learning fundamentals including supervised, unsupervised, and self-supervised learning, regularization, optimization, and model evaluation.
- Experience working with modern ML stacks including the Hugging Face ecosystem, experiment tracking (MLflow / W&B), and data versioning (DVC).
- Ability to design and execute rigorous ML experiments using scientific methodologies.
- Strong communication skills to present experimental findings and recommendations to both technical and non-technical stakeholders.
Nice to Have / Preferred
- Experience fine-tuning LLMs using PEFT techniques (LoRA, QLoRA, or Adapter-based methods).
- Knowledge of domain-specific AI applications including computer vision (object detection, segmentation), NLP (NER, sentiment analysis, summarization), or time-series forecasting.
- Experience with distributed training environments using multi-GPU infrastructure or cloud ML platforms.
- Familiarity with deployment optimization techniques including ONNX, TensorRT, and quantization approaches.
- Demonstrable portfolio such as published research papers (AI/ML conferences), open-source contributions, or competitive Kaggle projects.
- Master’s or PhD degree in Machine Learning, Data Science, Statistics, or Computer Science — highly preferred for this position.
- Experience in industries with complex AI requirements such as healthcare AI, financial risk modeling, or recommendation systems.
BINAR Academy aims to unlock and channel human potential so we can help the world turn into a better place. One of the ways we strive to do that is to channel potential talents to impactful opportunities using our very own job connector, BINAR Job Connect.
BINAR Job Connect has connected up to 250 talents & candidates; whether it be young & aspirational fresh graduate to tech-savvy team leaders, individuals who decided to start from scratch, and junior engineers to skilled product managers to our experienced hiring partners. BINAR Job Connect provides various options of employment, as we intend that everyone processed by us can choose the best career track for them.
