Tech Stack

Description

You will fine-tune, optimize, and deploy LLMs for multimodal social media content understanding across platforms like TikTok, YouTube, and Instagram. Your work spans from research and prototyping to production-grade serving at scale, building compound AI systems that combine multiple models into reliable pipelines.

Requirements

• Bachelor's/Master's in CS or related field with 4+ years ML/data science experience
• Hands-on experience fine-tuning large language models
• Experience with LLM inference optimization and serving frameworks (vLLM, TensorRT LLM)
• Fluency in Python and SQL (Snowflake)
• Experience with multimodal models and distributed GPU workloads

Responsibilities

• Fine-tune LLMs and perform inference optimizations (quality, latency, cost)
• Build compound AI systems combining multiple models and modalities
• Take ML models from research to production at scale
• Work with multi-terabytes of social media platform data
• Evaluate cost/quality tradeoffs across LLM providers and self-hosted models