LMTS - Lead AI/ML Engineer (LLM Post-Training)
Salesforce, Inc. (company site, opens in a new tab)10+ yrsBengaluruOn-siteDirect career pageEst. 45–85 LPA
Posted 6h ago
Reach people at Salesforce, Inc.
People at this company you can reach
Log in to see who works here
Create a free account or log in to see the people at Salesforce, Inc. you could write to:
- Recruiters & HR
- People in this function
- Work addresses
Your match
Against your primary resume
Log in to see your match score
Create a free account or log in to unlock your CV match score across:
- Role fit
- Experience
- Industry relevance
Protocol Intelligence
How competitive this role is
Competition
Background fit
Strictness
Log in to see why each signal reads the way it does.
Job description
Structured overview of role & requirements
About This Role
- Own the end-to-end post-training loop for open-weight LLMs: generate and curate data, run SFT/LoRA or justified preference/RL experiments, and release improvements backed by evaluation.
- Build and operate model evaluation, serving, and integrations with enterprise tools; balance model quality, latency, throughput, security, and reliability.
- Lead technical decisions and mentor engineers; diagnose failures across data, training, model behavior, tools, and backend systems.
Minimum Requirements
- Work Experience Required: 10+ years in software engineering, including technical leadership on production systems.
- Hands-on ownership of open-weight LLM post-training using SFT and LoRA/QLoRA; API integration alone does not qualify. Also requires hands-on training-data generation and curation.
- Practical post-training stack knowledge and experience evaluating models with held-out data, metrics, error analysis, experiments, and regression checks.
- Strong production Python/software-engineering fundamentals and demonstrated use of AI coding tools; advanced prompt engineering is also required.
Ideal Candidate Profile
- Has owned a model-training project from base model and data provenance through training, evaluation, and production outcome, and can explain personal contribution and failures with evidence.
- Can create reliable reasoning, tool-use, preference, or structured-output data and validate correctness, prevent contamination, and detect low-quality traces or reward hacking.
- Combines ML research engineering with production systems leadership; preference/RL methods, resource-constrained training or serving, and enterprise workflow experience are valued but not required.