Role Overview:

We are looking for a Senior AI/Machine Learning Engineer (7+ years of experience) who is a stellar individual contributor with recent hands-on expertise in fine-tuning and large-scale training of modern models (LLMs/VLMs).

This is a hands-on role where you will :

- Lead fine-tuning workflows, large-scale training runs, and evaluation design.

- Collaborate closely with researchers to bring cutting-edge approaches from papers into production.

- Work directly with customers to align performance metrics, validation, and deployment readiness.

Note: This is not a managerial role. We are seeking candidates who are currently active ICs, heavily involved in model fine-tuning/training in the last 8–12 months.

What You’ll Do Day-to-Day

- Fine-tune and train LLMs/VLMs at scale (LoRA/QLoRA, PEFT methods).

- Build reproducible training pipelines with distributed training and mixed precision.

- Design and run robust evaluation frameworks (task-specific + lm-eval-harness).

- Translate research papers into working implementations, collaborating with researchers.

- Work with customers to validate models against business and performance needs.

- Optimize training runs with profiling, performance tuning, and efficiency improvements.

- Maintain experiment tracking, reproducibility, and structured model artifacts.

Requirements

- 7+ years in ML/AI engineering, with a strong recent focus on fine-tuning and training large models.

- Expert in PyTorch, Hugging Face Transformers, and PEFT (LoRA/QLoRA).

- Strong experience with distributed training (DDP, FSDP, DeepSpeed, Accelerate).

- Skilled with evaluation frameworks (lm-eval-harness, custom metrics, task benchmarks).

- Proven ability to reproduce and improve results from recent research papers.

- Strong coding practices in Python, with modular, clean implementations.

- Familiarity with experiment tracking tools (Weights & Biases, MLflow).

- Ability to interact with customers and researchers to translate requirements into engineering solutions

[Bonus] Preferred Qualifications:

• Experience with other ML libraries (e.g., PyTorch, Flax)
• Background in ML research or scientific computing
• Experience with production model monitoring and governance
  

• للتقديم الان