Senior MLOps Engineer
Full Time - Remote @ITJobs.lk posted 6 days ago in AI & Machine Learning , in IT Infrastructure & Cloud Shortlist Email JobJob Description
Location: Fully Remote (Sri Lanka Based)
Type: Full-Time | Permanent
Timezone: Flexible (US or IST)
Industry: AI-Powered Ecommerce Infrastructure
The Opportunity
Our client, a leading IT firm in Sri Lanka, is seeking a Senior MLOps Engineer to join a high-impact team working —a cutting-edge US-based platform transforming ecommerce.AI-powered infrastructure turns static catalogs into intelligent product data for global brands.
You will own the deployment and optimization of model infrastructure, working directly with state-of-the-art open-source LLMs (QWEN, Llama, Mistral).
Key Responsibilities
-
Cost Optimization (Primary Focus): Architect strategies for inference cost reduction through batching, quantization, caching, and smart resource allocation.
-
Fine-Tuning Pipeline Management: Build and scale pipelines for open-source models (QWEN3, Llama, Mistral) using proprietary “golden datasets.”
-
Data Operations: Collaborate on the versioning and expansion of curated datasets; implement rigorous data quality and validation checkpoints.
-
Model Deployment: Maintain production ML pipelines capable of handling millions of SKU classifications with high reliability and low latency.
-
Infrastructure Architecture: Optimize ML environments on AWS (Bedrock, SageMaker, EC2/ECS) with future expansion into Azure; manage GPU clusters for training.
-
Performance Monitoring: Implement tracking for model drift, fine-tuning metrics, and system health to trigger retraining cycles.
-
Technical Leadership: Mentor the engineering team on MLOps best practices and drive critical architectural decisions.
Technical Requirements
-
Experience: 5+ years in MLOps, ML Engineering, or Platform Engineering.
-
LLM Expertise: Hands-on experience fine-tuning open-source LLMs (QWEN, Llama, Mistral) for production.
-
Dataset Management: Proven track record in managing training datasets, versioning, and pipeline automation.
-
Cloud (AWS Required): Deep expertise in Bedrock, SageMaker, Lambda, ECS/EKS, S3, and Step Functions.
-
Programming & Frameworks: Strong Python skills and mastery of PyTorch, Hugging Face Transformers, PEFT/LoRA, and DeepSpeed.
-
Optimization Techniques: Experience with LoRA, QLoRA, quantization (GPTQ, AWQ, GGUF), distillation, and VRAM optimization.
-
DevOps Tools: Expertise in Docker and Kubernetes.
- Communication: Exceptional English communication skills to collaborate effectively with US-based stakeholders.
Nice to Have
-
Experience with Azure ML or Azure OpenAI services.
-
Familiarity with inference frameworks like vLLM, TensorRT-LLM, or TGI.
-
Experience with multi-GPU/multi-node distributed training.
-
Knowledge of ML observability tools (Weights & Biases, MLflow, Langfuse).
What’s on Offer
-
Competitive Salary & Equity: Highly attractive compensation package.
-
Global Impact: Work on state-of-the-art AI technology at a massive scale.
-
Remote-First: Work from anywhere in Sri Lanka with flexible hours.
-
Career Growth: High-visibility role with direct influence on the technical roadmap.
How to Apply
If you are a Senior Engineer ready to lead the next generation of MLOps for global ecommerce, we want to hear from you.
Apply now via ITJobs.lk or send your CV to hello@itjobs.lk with the subject: Senior MLOps Engineer – [Your Name]
Other jobs you may like
-
Full-Stack AI Multimedia Developer Featured
- @ ITJobs.lk
- Sri Lanka