Machine Learning Engineer

Tranzeal; Bengaluru; On-site; INR 3,500,000 - 7,500,000; Full time

Job summary

Tranzeal is seeking a Senior LLMOps / MLOps Engineer to lead deployment and optimization of open-source LLMs at scale. The role focuses on building high-performance inference platforms using vLLM, SGLang, TGI, Triton, and Ray Serve, while ensuring GPU utilization, latency, throughput, and cost efficiency.

The ideal candidate will be hands-on, with 5-7 years in MLOps/AI platform engineering and strong experience with Kubernetes, Docker, Azure ML, Databricks, and MLflow.

Qualifications

• 5-7 years of experience in MLOps/LLMOps or AI platform engineering.
• Strong Python coding and software engineering practices.
• Experience with open-source LLMs (e.g., Llama, Mistral, Gemma, Qwen).
• Expertise in LLM inferencing and model hosting using vLLM, SGLang, TGI, Triton, Ray Serve, Azure ML, or Databricks Model Serving.
• Kubernetes, Docker, Azure ML, Databricks, and MLflow expertise.

Responsibilities

• Design, build, and operate production-grade LLM/GenAI platforms at scale.
• Optimize GPU utilization, latency, and throughput for inference workloads.
• Troubleshoot and improve end-to-end AI pipelines and deployment infra.
• Collaborate on governance, observability, and responsible AI practices for deployments.
• Explore and implement cutting-edge MLOps tooling and automation.

Skills

MLOps/LLMOps

Python

LLM Inferencing

LLM Hosting

GPU Optimization

Tools

Kubernetes

Docker

Azure ML

Databricks Model Serving

MLflow

Job description

Senior LLMOps / MLOps Engineer
Experience: 5-7 Location: ITPL, BLR
Work Mode: 5 days from office

Summary
We are looking for a highly skilled Senior LLMOps / MLOps Engineer with strong expertise in LLM inferencing, model hosting, and serving Large Language Models (LLMs) at scale. The ideal candidate should be a hands-on engineer with proven experience deploying and optimizing open-source LLMs, building high-performance inference platforms using technologies such as vLLM, SGLang, TGI, Triton, and Ray Serve, and driving GPU utilization, latency, throughput, and cost optimization. This is a highly technical role requiring active involvement in designing, building, troubleshooting, and optimizing production AI systems. Experience in MLOps platforms and scalable AI infrastructure is essential.
Must-Have Skills

• 5-7 years of experience in MLOps, LLMOps, AI/ML Platform Engineering.
• Strong proficiency in Python and software engineering best practices.
• Experience working with open-source LLMs such as Llama, Mistral, Gemma, or Qwen.
• Strong expertise in LLM Inferencing and Model Hosting using vLLM, SGLang, TGI, Triton, Ray Serve, Azure ML, or Databricks Model Serving.
• Experience with Kubernetes, Docker, Azure ML, Databricks, and MLflow.
• Good understanding of RAG, Vector Databases, GPU Optimization, Quantization, KV Cache, PagedAttention, and Continuous/Dynamic Batching.
• Demonstrated hands-on experience building, deploying, troubleshooting, and optimizing production-grade LLM and GenAI solutions.
• Experience deploying, scaling, and monitoring production-grade GenAI/LLM applications.
• Exposure to AI Observability, Governance, and Responsible AI practices.

Good-to-Have Skills

• Hands-on experience with LLM Fine-Tuning using PEFT, SFT, CPT, LoRA, and QLoRA techniques.
• Experience with Azure AI Foundry, Azure OpenAI, Hugging Face, DeepSpeed, and PEFT.
• Knowledge of distributed training and multi-GPU environments.
• Experience with Agentic AI frameworks such as LangGraph, AutoGen, or CrewAI.

Understanding of simulation platforms, digital twins, modeling & simulation workflows, or scientific computing