AI Research Scientist (Internship)

What you'll do

• Fine-tune vision-language models and curate training datasets from surveillance and security video.
• Run evaluation experiments to compare model performance across activity types and camera conditions.
• Support the pipeline from trained checkpoints to live-feed inference and summarize relevant VLM research.

What they ask for

• Coursework or project experience in machine learning, computer vision, or multimodal AI.
• Familiarity with PyTorch and at least one fine-tuning workflow, such as LoRA or SFT.
• Interest in vision-language models and applying them to real-world video.
• Curiosity, self-direction, and comfort working with messy, real-world data.

Tools

PyTorch, LoRA, SFT

AI focus

Fine-tune and evaluate vision-language models for real-time activity detection on live camera feeds.