Senior AI/ML Engineer – SLM Training & Inference Optimization

Concentrix; United States; Remote; USD 140,000 - 190,000; Full time

Job summary

Concentrix is seeking a Senior AI/ML Engineer to design, fine-tune, and optimize small language models for domain-specific, production-ready applications. This role focuses on model training, real-time inference optimization, and low-latency audio pipelines using modern AI/ML engineering practices.

Location is remote in the United States. You will work with Alibaba Qwen, PyTorch, and PEFT techniques to push for high-throughput, low-latency deployment and measurable improvements in accuracy and

Qualifications

• Strong hands-on experience with PyTorch, Hugging Face Transformers, and LLM/SLM fine-tuning techniques (LoRA, QLoRA, PEFT).
• Practical experience with vLLM or SGLang for high-performance inference.
• Deep understanding of GPU utilization, CUDA fundamentals, KV caching, and memory optimization.
• Experience evaluating and improving model precision, recall, and inference performance.
• Strong Python programming and ML engineering skills.
• Experience fine-tuning and deploying Alibaba Qwen models.
• Experience with real-time audio AI, speech recognition (ASR), or audio classification.
• Experience optimizing AI models for high-throughput, low-latency production environments.

Responsibilities

• Fine-tune and optimize Small Language Models (SLMs), particularly Alibaba Qwen models, for domain-specific applications.
• Improve model accuracy, precision, recall, latency, and inference throughput.
• Optimize model serving using vLLM, SGLang, or similar inference engines.
• Implement GPU memory optimization, KV-cache management, quantization, and efficient batching.
• Develop and optimize real-time audio intelligence pipelines for low-latency processing.
• Benchmark and tune models for production-scale deployment.

Skills

PyTorch

Hugging Face Transformers

LoRA / PEFT

vLLM / SGLang

CUDA / GPU optimization

Python

SLM fine-tuning

Alibaba Qwen

ASR / audio processing

Tools

vLLM

SGLang

Job description

We're Concentrix. The intelligent transformation partner. Solution-focused. Tech-powered. Intelligence-fueled.

The global technology and services leader that powers the world’s best brands, today and into the future. We’re solution-focused, tech-powered, intelligence-fueled. With unique data and insights, deep industry expertise, and advanced technology solutions, we’re the intelligent transformation partner that powers a world that works, helping companies become refreshingly simple to work, interact, and transact with. We shape new game-changing careers in over 70 countries, attracting the best talent.

The Concentrix Technical Products and Services team is the driving force behind Concentrix’s transformation, data, and technology services. We integrate world-class digital engineering, creativity, and a deep understanding of human behavior to find and unlock value through tech-powered and intelligence-fueled experiences. We combine human-centered design, powerful data, and strong tech to accelerate transformation at scale. You will be surrounded by the best in the world providing market leading technology and insights to modernize and simplify the customer experience. Within our professional services team, you will deliver strategic consulting, design, advisory services, market research, and contact center analytics that deliver insights to improve outcomes and value for our clients. Hence achieving our vision.

Our game-changers around the world have devoted their careers to ensuring every relationship is exceptional. And we’re proud to be recognized with awards such as "World's Best Workplaces," “Best Companies for Career Growth,” and “Best Company Culture,” year after year.

Join us and be part of this journey towards greater opportunities and brighter futures.

JOB SUMMARY

As a Senior AI/ML Engineer – SLM Training & Inference Optimization, you will help design, fine-tune, and optimize Small Language Models for domain-specific, production-ready applications. This role focuses on model training, real-time inference optimization, and low-latency audio intelligence pipelines using modern AI/ML engineering practices.

Location: [Remote- US]

Responsibilities

• Fine-tune and optimize Small Language Models (SLMs), particularly Alibaba Qwen models, for domain-specific applications.
• Improve model accuracy, precision, recall, latency, and inference throughput.
• Optimize model serving using vLLM, SGLang, or similar inference engines.
• Implement GPU memory optimization, KV-cache management, quantization, and efficient batching.
• Develop and optimize real-time audio intelligence pipelines for low-latency processing.
• Benchmark and tune models for production-scale deployment.

Qualifications

• Strong hands-on experience with PyTorch, Hugging Face Transformers, and LLM/SLM fine-tuning techniques (LoRA, QLoRA, PEFT).
• Practical experience with vLLM or SGLang for high-performance inference.
• Deep understanding of GPU utilization, CUDA fundamentals, KV caching, and memory optimization.
• Experience evaluating and improving model precision, recall, and inference performance.
• Strong Python programming and ML engineering skills.
• Experience fine-tuning and deploying Alibaba Qwen models.
• Experience with real-time audio AI, speech recognition (ASR), or audio classification.
• Experience optimizing AI models for high-throughput, low-latency production environments.

At Concentrix, we provide customer experience solutions that may involve handling sensitive data. As part of our hiring process, all candidates must undergo a background check in accordance with applicable law, which will include identity verification and employment eligibility.