Applied Scientist - AI/ML Engineer

We are looking for an Applied Scientist - AI/ML Engineer to join our team, focused on leveraging existing AI/ML models to identify key moments in sports videos through computer vision and transcript analysis. This role emphasizes prompt engineering, validation of model outputs, and integration with inference pipelines, rather than building or training custom models.

In this role, you will iterate extensively on Amazon Bedrock models to improve results by refining prompts that identify sports moments through transcripts and video frames within a multi-modal framework. You will run test video assets through Amazon Bedrock for custom moment detection, configure and tune models to boost accuracy, and create repeatable templates for pipeline processing and structured metadata output.

Responsibilities

• Design and refine prompts for multimodal foundation models (Claude, Nova, etc.) to identify custom ad moments in live video frames and audio transcripts
• Perform model comparison studies with accuracy benchmarking on domain-specific content
• Configure and adjust the Bedrock Connector pipeline for each custom moment type
• Verify detection accuracy across various sports, genres, and live event types
• Refine prompts iteratively to enhance detection accuracy and outcomes
• Partner with the AWS Elemental Inference team to align on technical direction
• Create Python scripts to automate custom moments pipelines when required
• Develop repeatable templates for pipeline processing and structured metadata output

Requirements

• 3+ years of experience in applied AI/ML with a background in multimodal models (vision + language)
• Hands-on expertise in prompt engineering for large foundation models, ideally using Amazon Bedrock
• Proficiency in Python for scripting and pipeline automation
• Familiarity with video/image understanding in live video workflows and near-real-time inference pipelines
• Capability to design evaluation frameworks and benchmark model accuracy
• Background in media, advertising, or content classification
• Understanding of Media Supply Chain concepts and workflows
• Proficiency in English at an Upper-Intermediate level (B2) or higher

Nice to have

• Flexibility to use OpenCV for handling video frames along with transcript data
• Knowledge of computer vision and natural language processing techniques
• Background in sports analytics or media technology