Research Scientist - Salesforce AI Research

Salesforce AI Research is looking for outstanding AI Research Scientists and Research Engineers to discover new research problems, develop novel models, and bridge the gap between groundbreaking theory and customer‑facing reality. Candidate Backgrounds: Agentic AI and Reasoning: 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. Compensation & Benefits: In the United States, compensation will be determined by factors such as location, job level, knowledge, skills and experience. The typical base salary range for this position is $117,200 - $313,700 annually. In select cities within the San Francisco and New York City metropolitan area, the base salary range is $141,200 - $344,700 annually. Benefits include: time off programs, medical, dental, vision, mental health support, paid parental leave, life and disability insurance, 401(k), and an employee stock purchasing program. Equal Opportunity Employment Salesforce is an equal opportunity employer and maintains a policy of non‑discrimination with all employees and applicants for employment. Any employee or potential employee will be assessed on the basis of merit, competence and qualifications – without regard to race, religion, color, national origin, sex, sexual orientation, gender expression or identity, transgender status, age, disability, veteran or marital status, political viewpoint, or other classifications protected by law. This policy applies to recruiting, hiring, job assignment, compensation, promotion, benefits, training, assessment of job performance, discipline, termination, and everything in between. Recruiting, hiring, and promotion decisions at Salesforce are fair and based on merit. #J-18808-Ljbffr