Job Description

Join our Team

About this opportunity

Ericsson’s Early Career Program is an 18-month journey for you who are recently graduated or have up to five years of relevant experience. Through three rotations, including one international and market exposure, you will tackle challenges that matter, work alongside people who expect you to contribute, question, and move things forward.

The Innovate track is designed for you who are driven by building what’s next. You’ll rotate across Research, Product Development, Data, and AI contributing to solutions that don’t just support the business but move it forward.

This isn’t a programme where you sit on the sidelines. From day one, you’re hired into a role, trusted to contribute, supported to grow, and expected to make an impact. Find out more about the Early Career program here.

Your journey within the Early Career Program will start as a LLM Post-Training Engineer at BNEW SAN AI Accelerator!

The AI Accelerator within BNEW SAN is transforming Customer Support and Network Deployment with agent-centric AI. As an LLM Post-Training Engineer, you’ll help adapt foundation models for telecom use cases through fine-tuning, alignment, and data work—making AI agents more reliable and useful while supporting the journey toward autonomous network operations.

What you will do

Working with experienced engineers and domain specialists, you’ll contribute to the areas below and grow your ownership over time.

Domain Knowledge Injection via Fine-Tuning

• Contribute to developing and running supervised fine-tuning (SFT) pipelines that adapt base LLMs with domain knowledge for Customer Support and Network Deployment use cases.
• Experiment with parameter-efficient fine-tuning techniques—including LoRA, QLoRA, and adapter-based methods—and help evaluate trade-offs in model quality, compute, and latency.
• Work with the AI & Domain Specialist and AI Data & Integration Engineer to review fine-tuning datasets for quality, domain coverage, and training-ready formats.
• Help document and organize fine-tuned model variants, adapter checkpoints, and training configurations to support reproducibility and model lifecycle management.

Policy-Reward Training & Alignment

• Support experiments with policy-reward training approaches, including methods such as GRPO and SAPO, with guidance from experienced teammates.
• Help prepare and evaluate reward signals and preference datasets that capture domain-specific quality—including correct, safe, and operationally appropriate behavior in telco contexts.
• Contribute to improving alignment strategies by reviewing evaluation results and production feedback with the team.
• Keep up with post-training research and share promising techniques for the team to assess.

Automatic Dataset Curation & Synthetic Data Generation

• Contribute to building and maintaining data-curation pipelines by helping source, filter, deduplicate, and format datasets for SFT and reward training.
• Help develop and evaluate synthetic-data approaches to create diverse training examples where real-world data is limited, sensitive, or imbalanced.
• Work with domain experts to review synthetic examples for quality and relevance to real telco operational scenarios.
• Help develop and run data-quality checks that maintain dataset integrity through curation and generation.

Training Run Management & Experiment Tracking

• Support training runs by helping configure jobs, schedule and monitor workloads, and manage GPU resources.
• Maintain clear experiment records with training configurations, dataset versions, evaluation results, and model artifacts to make results reproducible.
• Track and summarize key training metrics—such as loss curves, reward trends, benchmark scores, and alignment quality—for the AI System Architect and engineering team.
• Work with experienced teammates to investigate training instability, overfitting, and performance bottlenecks, and help test potential improvements.

The skills you bring

We welcome early-career candidates and do not require professional experience in LLMs or model training. We are especially interested in candidates who have explored how to train large language models through academic work, research, projects, internships, or self-directed learning.

Must-have

• Foundational knowledge of machine learning, data science, or natural language processing, and a strong interest in training large language models.
• Programming skills in Python, with a careful and curious approach to experimentation, data, and model evaluation.
• A collaborative mindset and the ability to communicate ideas and learn alongside technical and domain-specialist teammates.

Nice to have

• Academic or research experience focused on large language models or model training, including fine-tuning or post-training. Experience with SFT or parameter-efficient methods such as LoRA is a plus.
• Hands-on experience with LLMs through coursework, research, projects, or internships—even if not gained in a professional role.
• Familiarity with preference optimization or reinforcement learning approaches such as DPO, PPO, or GRPO or an interest in learning related approaches such as SAPO.
• Exposure to training-data curation, synthetic data, experiment tracking, model evaluation, or GPU-based training.
• Interest in telecommunications, RAN, or network operations.

Why join Ericsson?At Ericsson, you´ll have an outstanding opportunity. The chance to use your skills and imagination to push the boundaries of what´s possible. To build solutions never seen before to some of the world’s toughest problems. You´ll be challenged, but you won’t be alone. You´ll be joining a team of diverse innovators, all driven to go beyond the status quo to craft what comes next.

What happens once you apply?Click Here to find all you need to know about what our typical hiring process looks like.Encouraging a diverse and inclusive organization is core to our values at Ericsson, that's why we champion it in everything we do. We truly believe that by collaborating with people with different experiences we drive innovation, which is essential for our future growth. We encourage people from all backgrounds to apply and realize their full potential as part of our Ericsson team. Ericsson is proud to be an Equal Opportunity Employer. learn more.

Primary country and city: Japan (JP) || Yokohama

Req ID: 793035