Search
Go
This job is no longer accepting applications
This listing was closed on Jul 17, 2026.
Browse Open JobsForward Deployed Engineer - LLM Post-training
Summary by AIEngineer.careers
Forward Deployed Engineer - LLM Post-training is a AI Engineer role (full-time). with Reflection. in SAN FRANCISCO, US. Imported listing (source: linkedin.com). Apply on the employer's site (linkedin.com).
Imported listing (source: linkedin.com) · Apply on linkedin.com
SAN FRANCISCO · US
—
Full-time
MID
The sections below reproduce the third-party job description for reference. AIEngineer.careers does not write or control this text.
Imported job description
Sourced from linkedin.com
Reflection is seeking a Forward Deployed Engineer to join their Applied AI team, focusing on fine-tuning and deploying open-weight LLMs for enterprise customers. You will work hands-on with customer data, run fine-tuning workflows, build evaluation infrastructure, and deploy models to production. Collaborate with research teams and customers to adapt models to specific domains.
Responsibilities
- Fine-tune open-weight models for customer-specific use cases using SFT, preference optimization, and reinforcement fine-tuning.
- Build and maintain evaluation infrastructure: design eval suites, curate test sets, establish baselines.
- Prepare training data from raw customer inputs: inspect data quality, clean and format datasets.
- Debug and diagnose training and inference issues.
- Support end-to-end deployments across hybrid environments.
- Contribute to evolving playbooks and best practices.
Requirements
- Applied ML experience with hands-on fine-tuning of language models.
- Familiarity with SFT, DPO, RLHF or similar techniques.
- Understanding of evaluation methodology.
- Comfort with training infrastructure: GPUs, compute management.
- Strong software engineering fundamentals (Python).
Reflection
New York · US · 159+ employees
Reflection is an AI lab developing open-source foundation models and software agents for AI-assisted development. It positions itself as an open-model alternative to closed frontier AI labs.
