Staff Research Scientist, AI Agents & LLMs

At the intersection of autonomous agents and large language models (LLMs), you will build the models and systems that turn enterprise data into self-directed, decision-making agents—shifting from passive data access to autonomous execution.

Our team operates at the frontier of enterprise AI, delivering both state-of-the-art models and production systems, including:

Arctic LLM: Enterprise foundation models optimized for performance and efficiency

Arctic-Text2SQL: Post-trained reasoning models with frontier-level quality at a fraction of cost and latency

Snowflake Intelligence: Brought research to production through agentic innovations in multi-step reasoning, Deep Research for structured and unstructured insights, and system optimization

Arctic Inference: High-performance LLM serving stack (Shift Parallelism, SwiftKV), open-sourced and powering Snowflake Cortex

Arctic Long-Context Training: Enables 1M+ token context training on a single H200 GPU

Agent World Models (AWM): RL training across 1,000+ synthetic environments for tool-using agents

We focus on core challenges such as reliable reasoning, long-horizon planning, grounded generation, and scalable model and agent development.

IN THIS ROLE, YOU WILL:

Define research direction in Agentic AI and LLMs

Develop models: train, fine-tune, and align models for enterprise reasoning and tool use

Advance autonomous agents: drive state-of-the-art capabilities in multi-step reasoning and tool use

Advance the stack: retrieval, grounding, memory, and multi-agent coordination

Establish evaluation standards for reliability, safety, and efficiency

Drive research breakthroughs into production within Snowflake’s platform

Mentor and lead to amplify the team’s technical impact

Contribute to the field through publications and open source

WE’RE LOOKING FOR:

Ph.D. in Computer Science or a related field with a strong publication record

Expertise in LLM development (training, fine-tuning, alignment, or post-training)

Experience with agentic systems (multi-agent systems, tool use, or agent optimization)

Strong systems thinking under real-world constraints (latency, cost, reliability)

Proven technical leadership and end-to-end ownership

Ability to translate research into impactful systems

WHY SNOWFLAKE?

Frontier AI, Full Stack
Work across the full stack—from foundation models (Arctic LLM, Arctic Text2SQL) to agent systems (Snowflake Intelligence) and inference/training infrastructure (Arctic Inference, Arctic Training).

State-of-the-Art, Real Impact
Drive cutting-edge research that ships—powered by massive enterprise data and production workloads.

Hard, Meaningful Problems
Build autonomous systems for reliable reasoning over enterprise data—where correctness, cost, and latency all matter.

Open Research Culture
Publish, open-source, and help shape the field.

Snowflake is growing fast, and we are scaling our team to accelerate AI innovation.

Are you ready to build the brain for the Enterprise Data Cloud?

Every Snowflake employee is expected to follow the company’s confidentiality and security standards for handling sensitive data. Snowflake employees must abide by the company’s data security plan as an essential part of their duties. It is every employee's duty to keep customer information secure and confidential.

Snowflake is growing fast, and we’re scaling our team to help enable and accelerate our growth. We are looking for people who share our values, challenge ordinary thinking, and push the pace of innovation while building a future for themselves and Snowflake.

How do you want to make your impact?

For jobs located in the United States, please visit the job posting on the Snowflake Careers Site for salary and benefits information: careers.snowflake.com