Research Engineer, Post-Training (All Industry Levels)
Develops alignment algorithms, data pipelines, and sampling methods to optimize post-training AI models for performance and efficiency. Requires PhD or equivalent, ML expertise including reinforcement learning and transformers, and production code experience.
Responsibilities
- Develop alignment algorithms and loss functions to improve data sample efficiency.
- Write data pipelines to process diverse web data into a format models can ingest.
- Identify quality signals to understand our model’s performance in the real world.
- Design sampling algorithms to improve serving efficiency of large generative models.
Requirements
- At least PhD (or equivalent).
- Write clear and clean production-facing and training code.
- Experience working with GPUs (training, serving, debugging).
- Experience with data pipelines and data infrastructure.
- Strong understanding of modern machine learning techniques (reinforcement learning, transformers, etc).
- Track-record of exceptional research or creative applied ML projects.
Nice to Have
- Experience with product experimentation and A/B testing.
- Experience training large models in a distributed setting.
- Familiarity with ML deployment and orchestration (Kubernetes, Docker, cloud).
- Publications in relevant academic journals or conferences in the field of machine learning.
Senior Software Engineer, AI Platform
Senior Software Engineer building scalable AI infrastructure, agent orchestration frameworks, evaluation systems, and high-performance LLM serving at Mixpanel. Requires 5+ years experience and hands-on LLM/agent work.
Senior Machine Learning Systems Engineer
Build large-scale ML experimentation and training orchestration platforms, including agentic AI execution systems, to accelerate Ads ML development at Reddit. Requires 5+ years infrastructure experience and 2+ years building production ML platforms.
Staff Software Engineer, Agentic Platform
Senior individual contributor architecting and scaling agentic LLM systems that turn messy manufacturing data into reliable root-cause insights. Owns orchestration, retrieval, evaluation, and guardrails for non-deterministic production systems.