AI Engineer, LMTS

## Description

### Overview of the Role

Join an agile team with deep startup roots. We operate as a high-velocity 'startup-within-Salesforce,' following our recent acquisition. You’ll be managed by the same founders and engineers who built the original company, offering the autonomy of a small team backed by the global scale and trust of Salesforce. You will have the unique opportunity to build a '0 to 1' product with the founding team, while leveraging Salesforce's world-class Data Cloud and security infrastructure to drive immediate global impact at scale.

Global supply chains still rely on slow, manual processes—email, spreadsheets, and fragmented data. The economy moves fast but supply chains don’t, creating an inefficiency that affects the $13T of goods shipped annually and is one of the largest untapped opportunities in modern enterprise.

Agentforce Supply Chain is reimagining the supply chain with an AI-powered platform for designing, automating, and running end-to-end business processes, with seamless collaboration through familiar channels like email. For Salesforce, this represents a massive growth opportunity in the back office, with innovations that flow into the front office. Customers are clamoring for more, rapidly expanding their use cases as we enter an exhilarating growth phase. As one user put it: *“I’ve been waiting for this for 20 years.”*

### The Impact You Will Have

### As a Lead Applied Scientist for the Agentforce Supply Chain Foundational model team, you will be instrumental in the development of a novel foundational model that we are building for complex tabular datasets.

• Develop, implement, and deploy AI models to solve real-world complex problems in global supply chain and manufacturing.
• Apply and build upon state-of-the-art machine learning technologies, stay-up-to-date with current developments in the AI field, and use your expertise to inspire AI applications across the organization.
• Collaborate with cross-functional teams including Engineering, Design, Product Management, and industry experts to build high-quality product features that can be used by major companies around the world.
• Establish and scale robust evaluation frameworks by creating domain-specific benchmarks and automated testing suites; implement production telemetry pipelines to monitor real-world model performance, data drift, and system health.

### Requirements

• Masters/PhD in Computer Science or quantitative field with research in AI

• 5-7+ years of industry or post-Masters/PhD experience developing, building, and deploying Machine Learning models for real-world scenarios
• Strong experience in deep learning and graphs or network theory. Experience with at least one other form of machine learning, such as natural language processing, reinforcement learning, computer vision, LLMs, etc.
• Deep expertise in training large-scale models across distributed systems using frameworks like PyTorch Distributed (DDP/FSDP), DeepSpeed, or TorchTitan on GPU clusters.
• Proficiency with Python and PyTorch, Tensorflow, or JAX
• Demonstrated ability not only to use state-of-the-art machine learning techniques but also to innovate upon them.
• End-to-End MLOps Proficiency: Hands-on experience building and maintaining production-grade ML pipelines using tools such as Kubeflow, Airflow, or MLflow for experiment tracking, versioning, and automated retraining.

### Preferred Qualifications

• Experience with supply chain, manufacturing, or related problems

For roles in San Francisco and Los Angeles: Pursuant to the San Francisco Fair Chance Ordinance and the Los Angeles Fair Chance Initiative for Hiring, Salesforce will consider for employment qualified applicants with arrest and conviction records.