Drimva

Amazon

Sr. Conversational AI Product Design Technologist , Prime Video

$154K–$240K

Culver City, CA +4 · Senior · Full-time · Posted Oct 22, 2026

Location
Culver City, CA
Sunnyvale, CA
New York, NY
SEATTLE, WA
San Francisco, CA
Experience
5+ years of front-end technologist, engineer, or UX prototyper experience.
Education
Bachelor's or Master's degree in design, HCI, computer science, linguistics, or equivalent.

About the job

Shapes and prototypes Prime Video conversational AI experiences, translating customer needs into model behaviors and measurable, multimodal product experiences.

What you'll do

  • Build functional prototypes across voice, text, visual, and ambient surfaces, and help advance concepts toward implementation.
  • Write and refine system instructions and prompts; shape model behavior and build evaluations and conversational quality metrics.
  • Design recovery from ambiguity, misunderstanding, hallucination, and failure; collaborate with design, research, product, and engineering partners.

What they ask for

  • 5+ years of front-end technologist, engineer, or UX prototyper experience.
  • Available online portfolio with coding or prototyping samples demonstrating conversational AI work.
  • Hands-on experience prototyping with and designing for LLM-powered products.
  • Experience shaping AI systems through system instructions, prompts, and building or running evaluations.

Nice to have

  • Background in UX design or research, linguistics, computational linguistics, or IVR/voice design.
  • Experience with conversational quality metrics and LLM evaluations, multimodal prototyping, or accessibility standards.

Pay

154,200.00 - 208,600.00 USD annually (base salary); 177,300.00 - 239,900.00 USD annually (base salary); 169,600.00 - 229,500.00 USD annually (base salary)

Tools

LLMs, AWS Bedrock, reinforcement learning, LLM-as-a-judge, WCAG AA

AI focus

Prototypes and shapes generative AI conversational experiences, including prompts and system instructions, model behavior, evaluations, quality metrics, multimodal interactions, and graceful recovery.

Model-written summary · the original posting is authoritative

First seen Oct 23, 2026 · Last verified Nov 5, 2026 (UTC)