LLM Post-Train Eng. Early Career Program
Core
Adapt foundation models for telecom use cases through fine-tuning, alignment, and data work to make AI agents more reliable for Customer Support and Network Deployment.
Role type
Early-career LLM Post-Training Engineer
Builds
AI agents for autonomous network operations and customer support
Domain
Telecommunications / AI / Large Language Models
Deliverable
production ML models
Required skills
Python, supervised fine-tuning (SFT), parameter-efficient fine-tuning (LoRA, QLoRA), policy-reward training (GRPO, SAPO), dataset curation, synthetic data generation, experiment tracking, GPU resource management
Preferred skills
Academic/research experience in LLMs, preference optimization (DPO, PPO), reinforcement learning, telecommunications domain knowledge
Technologies
LoRA, QLoRA, GRPO, SAPO, DPO, PPO, Python, GPU
Responsibilities
Develop and run SFT pipelines to adapt base LLMs with domain knowledge; Experiment with parameter-efficient fine-tuning techniques and evaluate trade-offs; Support experiments with policy-reward training approaches; Build and maintain data-curation pipelines for SFT and reward training; Manage training runs, schedule workloads, and monitor GPU resources; Track and summarize key training metrics for the engineering team.