Open positions
Research Program Manager Model Development
A research program manager who embeds with pre-training ML and data teams at an AI model developer, coordinating experiment planning, run scheduling, data readiness, checkpoint handoffs, and research milestones across training and infrastructure teams, while advising research leadership on technical tradeoffs.
Member of Technical Staff Safety
Owns red-teaming and adversarial evaluation pipelines for Reflection's AI models — probing for security, misuse, and alignment failures, implementing jailbreaks and defenses, building automated safety benchmarks, and validating model releases against risk thresholds. Core focus: LLM safety and ML evaluation systems.
Member of Technical Staff - Pre-Training
Research-focused engineer on Reflection's pre-training team: designs and runs experiments on language-model scaling, data efficiency, optimizers, and model architecture; implements deep learning methods, leads small research projects, and optimizes distributed training infrastructure, primarily using Python and PyTorch.
Member of Technical Staff - Pre-Training Infra
Builds and operates distributed training infrastructure for frontier-model pre-training across thousands of GPUs, optimizing throughput, stability, memory, and GPU utilization while maintaining dataset/checkpointing pipelines and debugging bottlenecks. Core tech includes Megatron, DeepSpeed, NCCL, and model parallelism.
Member of Technical Staff - Data Quality Engineer Pre-training
Reflection is hiring a Data Quality Engineer for LLM pre-training to define measurable data-quality standards, build automated QA pipelines and human-in-the-loop checks, and monitor quality trends using Python and ML/LLM workflows. On-site roles in San Francisco, New York, or London.
Member of Technical Staff - Data Quality Engineer Post-training
Reflection AI is hiring a Data Quality Engineer for its post-training team. Day to day, this person defines measurable data-quality standards for LLM training and evaluation, builds automated QA pipelines in Python over large datasets, gives feedback to external data vendors, and monitors quality trends to improve acceptance criteria.
Member of Technical Staff Post Training
A research-engineering role on Reflection's post-training team: you build systems that turn pre-trained LLMs into aligned, general agents, developing data-generation pipelines, reward models, reinforcement-learning algorithms, and inference-time scaling techniques while collaborating across pre-training and post-training efforts.
Member of Technical Staff Data Flywheel
Member of Technical Staff on Reflection's Data Flywheel team (New York, London, or San Francisco) who sources high-value data, designs evaluations, graders, and reward signals, analyzes LLM failure modes, and builds data pipelines that turn human and synthetic data programs into measurable model improvements.
Forward Deployed Engineer AI Engineer
You will architect, build, and deploy agentic AI systems for enterprise customers. You will orchestrate LLM workflows, integrate enterprise infrastructure, fine-tune models, and support reliable deployments across…
Lightspeed