The Experience
Salesforce AI Research is looking for outstanding AI Research Scientists and Research Engineers. Our team discovers new research problems, develops novel models, and bridges the gap between groundbreaking theory and customer-facing reality. We are dedicated to a high-impact research model that provides the best of both worlds: the ability to advance the state-of-the-art in AI and the opportunity to see that work operate at massive, real-world scale within the world's #1 AI CRM. We believe that making substantive progress on hard, enterprise-level applications drives and sharpens the research questions we study — and that the best researchers are also builders.
Learn more about our team at salesforceairesearch.com.
Candidates have a strong background in one or more of the following areas:
- Agentic AI and Reasoning: Large language model (LLM)-powered agents, reinforcement learning (RL), reasoning and planning, autonomous workflows, self-evolving agent systems, computer use agents, deep research agents, and coding agents.
- Multimodal and Computer Vision: Vision-language models (VLM), video understanding, visual grounding for agents, and multimodal AI agents.
- Speech Intelligence: Voice intelligence, text-to-speech/automatic speech recognition (TTS/ASR), human-like turn-taking, and low-latency audio processing.
- Core Modeling and Post-Training: Machine learning methodology, pre-training and post-training techniques including reinforcement learning from human feedback (RLHF) and direct preference optimization (DPO), model distillation, evaluation, causal inference, and time series modeling.
What You'll Actually Be Doing
- Discover new research problems, develop novel models, and translate cutting-edge ideas into practical, high-performance implementations that reach millions of users.
- Design, train, evaluate, and deploy neural networks and agent systems for enterprise applications at massive scale.
- Lead LLM fine-tuning, post-training, and model optimization tasks, including the application of RL and simulation to improve model accuracy and reliability.
- Collaborate across a cross-functional team of researchers and engineers to deliver complex projects that directly impact our core AI products and customer experiences.
You're Our Person If...
- You have a Ph.D., Master's, or equivalent industry/research experience in Computer Science, Machine Learning, or a related technical field.
- You have a strong publication record at top AI venues (e.g., NeurIPS, ICLR, ICML, CVPR, ICCV, ACL, EMNLP) and/or a proven track record of shipping AI research into real-world products.
- You are motivated by real-world impact — you care about seeing your research leave the lab and reach users at scale.
- You bring demonstrated expertise in one or more core focus areas: LLMs, Agentic Systems, RL, Computer Vision and Multimodal, or Speech Intelligence.
- You are proficient in Python and deep learning libraries such as PyTorch, JAX, or DeepSpeed.
Even Better If...
- You have experience training and evaluating large-scale multimodal or speech models, including techniques for alignment, safety, and latency optimization.
- You have designed and deployed AI agents — including computer use agents, deep research agents, or tool-using agents — in real-world or research settings.
- You have hands-on experience with RL, simulation, or self-evolving agent systems.
- You have experience with production cloud deployments (AWS/GCP), container orchestration (Kubernetes), or machine learning operations (MLOps).
- You are thoughtful about responsible AI, trust, and the real-world impact of the systems you build.