Senior Applied Scientist

NinjaTech AI; Sydney; On-site; AUD 180,000 - 230,000; Full time

Job summary

NinjaTech AI is seeking a Senior Applied Scientist to drive production-scale AI systems, including RL, LLMs, and multimodal capabilities. You will own end-to-end model development, from research ideas to scalable implementations, collaborating with engineers and product teams in Sydney, Australia.

You will advance reinforcement learning techniques, train large models, refine datasets, and optimize latency, throughput, and cost while shipping robust production components in Python and Rust.

Qualifications

• Master's degree or PhD in Computer Science, ML, or related quantitative field.
• Significant experience building end-to-end ML or DL systems from data to production.
• Deep understanding of ML, DL, probability, statistics, optimization, and algorithms.
• Hands-on experience with reinforcement learning methods (PPO, actor-critic, RLHF).
• Experience training and deploying transformer-based language models.
• Strong Python skills; Rust experience is valued.
• Deep experience with PyTorch and modern ML tooling.

Responsibilities

• Research, prototype, train, evaluate, and deploy advanced ML/LLM systems with focus on RL and agentic AI.

Skills

Python

PyTorch

Reinforcement learning

Transformer models

Distributed training

Multi-GPU training

Model optimization

NLP/NLG

Data pipelines

RLHF

Education

Master's degree in Computer Science/ML

Tools

ONNX

CUDA

Rust tooling

Job description

Come build the intelligence behind autonomous AI employees.

At NinjaTech AI, we're building AI systems that do more than answer questions. Our agents plan, reason, use tools, operate computers, collaborate with humans, and execute complex work over long periods of time. As a Senior Applied Scientist, you'll help push the capabilities of these systems forward through applied research in reinforcement learning, LLM training, model optimization, natural language systems, and multimodal AI.

This is a highly hands-on role for someone who wants to take cutting-edge research and make it work at production scale. You'll work closely with our Chief Science Officer, engineering and product leaders, ML engineers, and software developers, with significant ownership over the models and systems that power Ninja.

What You'll Own

• Research, prototype, train, evaluate, and deploy advanced ML and LLM systems, with a particular focus on reinforcement learning and agentic AI.
• Develop and apply state-of-the-art RL techniques, including RLHF and related approaches, to improve model reasoning, reliability, accuracy, and task performance.
• Design large-scale training and evaluation datasets, including synthetic data generation, statistical modeling, and data-mining techniques.
• Train, fine-tune, distill, quantize, and optimize large language models for production environments.
• Build high-performance multi-GPU training and inference pipelines using modern acceleration and mixed-precision techniques.
• Develop rigorous evaluation frameworks and experimentation systems to measure model quality, behavior, and performance.
• Design and run experiments, including A/B tests, to validate model and product improvements.
• Build robust, scalable production ML systems and write high-quality code primarily in Python and Rust.
• Optimize algorithms and systems for latency, throughput, memory efficiency, and cost.
• Advance NinjaTech's natural-language and speech capabilities, including NLU/NLG, automatic speech recognition, TTS, and STT.
• Translate novel academic research into practical capabilities that can be shipped to users.
• Continuously evaluate emerging research, models, and techniques and determine where they can meaningfully improve Ninja's AI systems.

What We’re Looking For

• Master's degree or PhD in Computer Science, Machine Learning, Applied Mathematics, Statistics, Engineering, or a related quantitative field.
• Significant experience building end-to-end machine learning or deep learning systems, from data and experimentation through training and production deployment.
• Deep understanding of machine learning, deep learning, probability, statistics, optimization, and algorithms.
• Strong experience with reinforcement learning methods such as PPO, actor-critic methods, RLHF, or related approaches.
• Hands-on experience training, fine-tuning, evaluating, and deploying transformer-based language models.
• Strong proficiency in Python; experience with Rust is highly valued.
• Deep experience with PyTorch and modern ML tooling and libraries.
• Experience with distributed and multi-GPU training, mixed precision, model distillation, quantization, ONNX, and inference optimization.
• Experience in one or more of the following areas: natural language understanding, natural language generation, information retrieval, question answering, knowledge extraction, text classification, or speech/audio processing.
• Strong understanding of software architecture, data structures, algorithms, database systems, and runtime/performance analysis.
• Ability to move comfortably between research and engineering, from exploring a new technique in a paper to shipping a reliable production implementation.
• Strong analytical instincts and the ability to independently investigate ambiguous problems, form hypotheses, run experiments, and make evidence-based decisions.

At NinjaTech AI, you won't be optimizing a small component of a mature system. You'll help shape the core intelligence of an autonomous AI platform and turn advances in AI research into capabilities used by real customers.