AI Research Engineer (Multi-Modal & Vision)

Tasks

• Apply knowledge distillation
• Apply reinforcement learning from human feedback
• Benchmark model quality and robustness
• Build multimodal datasets
• Build supervised fine tuning pipelines
• Design post training methodologies
• Develop evaluation frameworks
• Develop vision-language models
• Monitor and resolve training performance bottlenecks
• Optimize models for deployment
• Publish research results
• Run distributed training on GPU infrastructure
• Train and evaluate models

Perks/Benefits

• Autonomy
• Flexible working arrangements
• Fully remote
• Global collaboration
• Publication support

Skills/Tech-stack

Benchmarking | Distributed Training | Efficient Fine Tuning | Evaluation Frameworks | Fine Tuning | GPU infrastructure | GitHub | Hugging Face | Human Feedback | Knowledge Distillation | Learning from Human Feedback | Machine Learning | Machine Learning Pipelines | Model Compression | Model Optimization | Multimodal Learning | Open Source | Open-source AI | Parameter efficient fine-tuning | Reinforcement Learning | Reinforcement Learning from Human Feedback | Supervised Fine Tuning | Vision-language