Location: Lahore, Karachi, Islamabad
About Liquid Technologies
Liquid Technologies is a future-focused technology company delivering AI, software development, data engineering, DevOps, product design, and digital transformation solutions. Our multidisciplinary team of 100+ specialists has spent over a decade helping startups, enterprises, and organizations across healthcare, energy, manufacturing, finance, and beyond turn bold ideas into scalable, real-world products — from proprietary platforms like Vidan.AI (AI-powered video analytics) and Vitalog (digital health) to custom enterprise SaaS and machine learning pipelines. We pair deep technical expertise with a design-thinking approach and a strong commitment to responsible, ethical AI — engineering impact, not just code.
Role
Liquid Technologies is seeking an MLOps Engineer – ML Platform & Feature Store to build, operate, and scale core components of our machine learning platform. This role is ideal for a hands-on engineer who thrives in production ML environments, working closely with data scientists to enable reliable model training, evaluation, and deployment workflows. You will be responsible for managing feature pipelines, training jobs, MLflow operations, and evaluation systems, while ensuring platform stability, scalability, and reproducibility. This is a highly collaborative role working alongside Data Scientists and MLOps leadership to evolve the ML platform.
Responsibilities
- Build and maintain feature pipelines using Feast, including feature definitions and materialisation jobs (batch + streaming).
- Develop and manage training pipelines, including containerization, scheduling, dataset access, and artifact handling.
- Operate and maintain the MLflow tracking server, managing experiments, models, and artifact storage.
- Execute model evaluation workflows, run evaluation suites, and support model promotion decisions.
- Enable data scientists by resolving issues related to environment setup, data access, compute, and reproducibility.
- Manage GPU-based workloads and ensure efficient scheduling and utilization.
- Support distributed data processing using Spark or similar frameworks.
- Ensure air-gap readiness by managing dependencies, pre-building images, and enabling offline deployments.
- Collaborate with the MLOps Lead on platform improvements, scalability, and long-term architecture.
Requirements
- Bachelor’s degree in Computer Science, Engineering, or a related field (preferred).
- 3–5 years of experience in ML engineering, data engineering, or MLOps roles.
- Strong Python skills with experience in pandas, numpy, pyarrow, scikit-learn.
- Hands-on experience with feature stores (Feast preferred) or similar feature pipeline systems.
- Experience with MLflow or similar experiment tracking/model registry tools.
- Familiarity with distributed computing frameworks (Spark or equivalent).
- Working knowledge of Docker, Kubernetes (kubectl, Helm), and containerized workflows.
- Experience handling GPU-based workloads.
- Strong problem-solving skills and ability to support cross-functional teams.
Nice to Have
- Experience scaling Feast materialisation pipelines.
- Familiarity with Kubeflow, Tekton, or pipeline orchestration tools.
- Exposure to LLM workflows (vLLM, TGI, fine-tuning).
- Experience with data versioning tools (DVC, LakeFS, Nessie).
- Familiarity with evaluation frameworks (RAGAS, DeepEval).
