Skip to content
DatabricksDatabricksMountain View, CA

Principal Research Scientist - AI Scaling & Optimization

Leads research team advancing LLM scaling, efficiency, post-training, RL, and inference optimization. Drives innovations from research to production, requiring deep ML expertise, Python/PyTorch proficiency, and leadership in large-scale experiments.

270k – 350k
On-siteAI Research

About the role

Responsibilities

  • Lead and grow a multidisciplinary research team focused on LLM scaling, efficiency, and systems performance.
  • Define scaling research roadmap aligned with strategic objectives, prioritizing foundation model efficiency and large-scale training/inference.
  • Drive algorithmic innovations for neural network training/inference (optimizers, low-precision techniques, model adaptation) with empirical validation.
  • Optimize end-to-end ML systems for distributed training, RL, memory/compute efficiency in collaboration with systems/platform teams.
  • Partner with product/engineering to productionize research breakthroughs in scaling/efficiency.
  • Represent research externally via publications, talks, collaborations.
  • Mentor talent with technical/career guidance.

What You Will Do

  • Define/lead research programs on foundation model efficiency (optimizer design, low-precision training/inference, scalable architectures, adaptation methods).
  • Oversee large-scale experiments, benchmarking, trade-off evaluation (quality, latency, throughput, cost).
  • Write high-quality Python/PyTorch code for implementation, prototyping, production integration.
  • Collaborate on distributed training, parallelism, memory management, hardware utilization.
  • Establish metrics/evaluation protocols for scaling research (training efficiency, inference cost, energy usage).
  • Champion responsible deployment ensuring model reliability/safety.

Requirements

  • Proven leadership developing novel foundation model efficiency techniques with industry impact.
  • Deep expertise in generative AI, LLMs, distributed ML systems, model optimization, or responsible AI (focus on scaling/efficiency).
  • Strong programming in Python and PyTorch for research/implementation.
  • Ability to translate research to scalable product capabilities.
  • Excellent communication, leadership, stakeholder management.

Nice to Have

  • Experience at systems/ML intersection (distributed training frameworks, compiler/kernel optimization, memory/compute-efficient design).
  • Strong network in large-scale ML with conference service/publications (ICLR, ICML, NeurIPS, MLSys).
  • Record of research impact via publications, open-source, deployed systems in optimization/efficiency.

Skills

PyTorchPythonLLMsDistributed TrainingModel OptimizationLow-Precision TrainingRlNeural NetworksScalable ArchitecturesInference Optimization

Similar roles

AI Research jobs
Databricks

Principal Research Scientist – Scaling

DatabricksSan Francisco, CA

Leads research team advancing LLM scaling, post-training, RL, and inference efficiency. Drives innovations in optimization, distributed systems, and production integration using Python/PyTorch, with deep expertise in large-scale ML.

280k – 350k
HybridAI Research
Airbnb

Principal AI/ML Researcher / Engineer Reasoning, Planning, and Decision-making systems

AirbnbUnited States

Seeking a Principal AI/ML Researcher/Engineer to invent, scale, and operationalize intelligent decisioning systems. This role focuses on integrating Large Reasoning Models with Knowledge Graphs and applying Reinforcement Learning for adaptive planning and control in multi-agent systems.

296k – 370k
Remote15+ YOEAI Research
Datadog

Distinguished Architect, AI

DatadogNew York, NY +1

As a Distinguished Architect, AI, you will be a technical multiplier for leading AI labs and AI-native companies, bridging their infrastructure aspirations with Datadog's technology roadmap. You will ensure the platform solves unique observability challenges for training and deploying foundational models at scale.

300k – 484k
Hybrid10+ YOEAI Research
Upstart

Principal Applied Scientist

UpstartUnited States

As a Principal Applied Scientist, you will define the technical direction for offer optimization and conversion modeling systems, working across teams to integrate models and optimization systems. This role involves structuring ambiguous problems, designing solutions, and providing technical oversight to ensure a coherent long-term vision.

220k – 330k
Remote8+ YOEAI Research
MongoDB

Principal Research Scientist, Database Systems

MongoDBNew York, NY +1

Leads research in database systems, focusing on query optimization, storage engines, indexing, and hardware-accelerated architectures. Collaborates with engineering to productize innovations and builds a research team. Requires PhD and 12+ years experience.

168k – 330k
Remote12+ YOEAI Research