AI Research, Salesforce

• We are looking for entrepreneurial Research Interns for Summer 2027 who want to help build the next generation of Enterprise AI. This is not a traditional research internship
• We are looking for people who are comfortable with ambiguity, bring energy and focus to their work, and care as much about customer adoption as technical innovation
• You have a strong publication record at top AI venues (e.g., NeurIPS, ICLR, ICML, CVPR, ICCV, ACL, EMNLP) and/or experience building AI research into real-world systems
• You have experience with Python and deep learning frameworks such as PyTorch, JAX, or DeepSpeed
• You are motivated by real-world impact and care about seeing your research leave the lab and reach users at scale
• You bring expertise in one or more core focus areas:
• Agentic AI and Reasoning: LLM-powered agents, reinforcement learning (RL), reasoning and planning, autonomous workflows, computer use agents, deep research agents, and coding agents
• Multimodal and Computer Vision: Vision-language models, video understanding, and multimodal AI agents
• Speech Intelligence: Voice intelligence, text-to-speech/automatic speech recognition, and low-latency audio processing
• Core Modeling and Post-Training: Pre-training and post-training techniques including RLHF and direct preference optimization, model distillation, evaluation, and time series modeling
• You can conduct research with guidance from mentors, and you are ready to grow toward independence and translate ideas into impactful systems

Desirable

• Salesforce AI Research is a global leader in Enterprise AI, driving innovations that advance the state of AI while powering enterprise products at worldwide scale
• Our research has helped shape multiple generations of modern AI: we helped pioneer prompt engineering through DecaNLP; introduced BLIP, a foundational breakthrough in multimodal AI; trained state-of-the-art large language models including CodeGen; and pioneered foundation models for time series with Moirai
• Today, we continue to advance the frontier of enterprise AI through innovations in AI simulation environments like eVerse, our simulation and agent-training environment; the Atlas Reasoning Engine, which powers Agentforce reasoning capabilities; and other AI systems that are redefining how enterprises reason, plan, and act
• What makes our team unique is our ability to take ideas from research to real-world impact at scale. We work closely with product teams and enterprise customers to translate cutting-edge research into production AI capabilities used by millions of users. We also continue to publish at the world's top AI conferences, open source influential research, and collaborate with leading academic institutions
• You will work like a founder within Salesforce: exploring high-impact research directions, prototyping cutting-edge AI capabilities, and seeing how ideas are validated with customers and turned into production-ready solutions
• You will have the freedom to propose and explore new ideas, alongside mentors and research scientists, with access to Salesforce's global customer and product ecosystem
• If you want the pace and ownership of a startup, combined with the reach and resources of Salesforce, this internship offers a unique opportunity to do research with global impact
• Collaborate with research scientists and engineers on a project focused on real-world product impact, with the opportunity to publish your work at top-tier conferences
• Explore research directions with the potential to transform Enterprise AI
• Help turn promising ideas into zero-to-one AI product prototypes
• Build models, agents, and AI systems for real-world enterprise use
• Learn how research partners with product, engineering, and go-to-market teams to bring solutions to market
• This is a Summer 2027 internship with a minimum duration of 12 weeks