Principal Data Scientist

Salesforce•Palo Alto, CA

About The Position

About Salesforce Salesforce is the #1 AI CRM, where humans with agents drive customer success together. Here, ambition meets action. Tech meets trust. And innovation isn’t a buzzword — it’s a way of life. The world of work as we know it is changing and we're looking for Trailblazers who are passionate about bettering business and the world through AI, driving innovation, and keeping Salesforce's core values at the heart of it all. Ready to level-up your career at the company leading workforce transformation in the agentic era? You’re in the right place! Agentforce is the future of AI, and you are the future of Salesforce. The AgentForce Data Science team powers the core Large Language Models (LLMs) behind Salesforce’s production-grade AI agents. Our work directly impacts millions of users by enabling trustworthy, scalable, and high-performance AI systems across customer support, sales, marketing, analytics, and internal productivity workflows. We operate at the intersection of cutting-edge research and real-world deployment, owning the full model development lifecycle—from research ideation and training to fine-tuning, evaluation, continuous learning, and production rollout. Role Overview We are seeking a strong Lead/Principal Applied Scientist to drive advanced LLM research and model development for AgentForce’s production services. This role requires hands-on involvement across the full model development lifecycle, in addition to technical leadership and mentorship. The ideal candidate is both a strong individual contributor and a technical leader, serving as a primary point of contact (POC) for major AI initiatives while shaping long-term research and modeling strategy.

Requirements

PhD in Computer Science, Machine Learning, AI, or a related field.
Strong publication record in top-tier venues (e.g., NeurIPS, ICML, ICLR, ACL, EMNLP) or equivalent industry research impact.
Demonstrated hands-on experience owning the full model development lifecycle, not limited to research or design.
Deep expertise in large-scale model training and fine-tuning, especially for LLMs.
Strong background in reinforcement learning, preference learning, or human-in-the-loop learning.
Experience building and maintaining continuous learning systems using real-world feedback signals.
Solid understanding of model evaluation, alignment, and robustness in production environments.
Advanced proficiency in Python, with significant hands-on coding experience.
Deep experience with PyTorch, TensorFlow or similar deep learning packages.
Practical experience with modern LLM tooling, such as: Hugging Face (Transformers, Accelerate, PEFT) Distributed training frameworks (DeepSpeed, FSDP, etc.) ML orchestration and scaling tools (Ray, Kubernetes, internal platforms)
Strong data analysis and experimentation skills (NumPy, Pandas, custom evaluation pipelines).
Experience mentoring and developing junior researchers or engineers.
Strong communication skills across research, engineering, and executive stakeholders

Nice To Haves

Experience deploying and iterating on models in production, high-availability systems.
Background in enterprise AI, agentic systems, or LLM platforms at scale.
Familiarity with trust, safety, or governance frameworks for AI systems.
Experience with large-scale distributed compute environments (multi-GPU / multi-node training).

Responsibilities

Own and execute hands-on work across the full model development lifecycle, including data preparation, model training, fine-tuning, evaluation, iteration, and deployment readiness.
Lead end-to-end research initiatives on LLM training, fine-tuning, alignment, and optimization for production use cases.
Design, implement, and iterate on reinforcement learning (RL) and continuous learning pipelines (e.g., RLHF, RLAIF, offline/online feedback loops).
Conduct rigorous experimentation, ablation studies, and failure analysis to drive measurable model improvements.
Translate research prototypes into production-grade models that meet latency, scalability, reliability, and safety requirements.
Serve as the technical POC for complex AgentForce AI projects, driving alignment across research, engineering, product, and platform teams.
Define best practices for model training, fine-tuning, evaluation, and release readiness.
Influence architectural and modeling decisions across the AgentForce AI stack.
Mentor junior scientists and engineers through direct technical guidance and code-level reviews.
Foster a culture of strong scientific rigor, reproducibility, and ownership.
Contribute to Salesforce’s external research presence through publications, talks, and collaborations.