Founding Engineer - Post-training LLM's

CT19; New York (NY); On-site; USD 180,000 - 260,000; Full time

Job summary

CT19 is partnering with an early-stage startup at the intersection of BioTech and AI, building a core intelligence layer for biology. We are seeking a Founding Research Engineer to help build and scale the systems that power the company’s models.

This role sits at the intersection of research and engineering, focusing on training, post-training, evaluation, performance optimisation, and the infrastructure required to support these processes.

Qualifications

• Experience training or significantly improving advanced LLMs or generative ML systems.
• Strong software engineering and distributed systems expertise.
• Proficiency in Python and ML frameworks such as PyTorch or JAX.
• Experience with reinforcement learning or post-training methodologies.
• Background building evaluation systems for tool-using or open-ended models.
• Understanding of GPU memory constraints and performance trade-offs.
• Experience diagnosing performance issues in production ML environments.
• Comfortable balancing research exploration with engineering execution.

Responsibilities

• Build and improve training and post-training systems for biological foundation models and agent-based workflows.
• Design and run experiments across supervised fine-tuning, reinforcement learning, tool use, evaluation, and model behaviour optimization.
• Develop and maintain distributed RL and post-training infrastructure.
• Improve reliability of rollout, evaluation, and reward pipelines.
• Identify and resolve performance bottlenecks across GPU, networking, and storage layers.
• Collaborate with founders and domain experts to translate biological problems into model tasks and evaluation frameworks.
• Contribute to translating research advancements into tangible product and customer impact.

Skills

LLM training

Distributed systems

Python

PyTorch/JAX

Reinforcement learning

Evaluation systems

GPU performance

Production ML

Research-to-implementation

Tools

PyTorch

JAX

Job description

CT19 are partnering with an early stage start up operating at the intersection of BioTech and AI, who are building a core intelligence layer for biology.

Their technology has been developed through leading academic and industry collaborations and published in top machine learning venues and within a few months of spinning out, their models are already being adopted by top 10 pharma companies, on challenges such as response prediction, biomarker discovery, and scientific reasoning across complex biological datasets.

About the Role

We are supporting the search for a Founding Research Engineer to help build and scale the systems that power the company’s models.

This position sits at the intersection of research and engineering, focusing on training, post-training, evaluation, performance optimisation, and the infrastructure required to support these processes.

Responsibilities

• Build and improve training and post-training systems for biological foundation models and agent-based workflows
• Design and run experiments across supervised fine-tuning, reinforcement learning, tool use, evaluation, and model behaviour optimisation
• Develop and maintain distributed RL and post-training infrastructure
• Improve reliability of rollout, evaluation, and reward pipelines
• Identify and resolve performance bottlenecks across GPU, networking, and storage layers
• Collaborate with founders and domain experts to translate biological problems into model tasks and evaluation frameworks
• Contribute to translating research advancements into tangible product and customer impact

Candidate Profile

• Proven experience training or significantly improving advanced LLMs or generative ML systems
• Strong software engineering and distributed systems expertise
• Deep proficiency in Python and modern ML frameworks such as PyTorch, JAX, or similar
• Experience with reinforcement learning or post-training methodologies
• Background in building evaluation systems for tool-using or open-ended models
• Strong understanding of GPU performance constraints and memory trade-offs
• Experience diagnosing and resolving performance issues in production ML environments
• Comfortable balancing research exploration with engineering execution
• Comfortable operating in a fast-moving, early-stage environment