ML Research Engineer - Workflows/Systems

Summary

Achira is a company focused on innovative drug discovery through advanced machine learning and AI. They are seeking an ML Research Engineer to design and implement machine learning systems and workflows that enable scientific experimentation at scale.

Responsibilities

• Build and maintain robust multi-stage asynchronous workflows for running data generation, training, and evaluations for our machine learning stack
• Rationalize machine learning systems design and software architecture
• Identify blockers and build solutions that scale to the size of foundation models
• Operate as the glue between research scientists and the infrastructure team

Skills

• At least two years relevant industry experience
• Highly fluent in and enthusiastic about PyTorch and JAX
• Used to thinking in asynchronous primitives
• Strong views on library design: clean abstractions, minimal surface area, consistency
• Solid track record of observable artifacts (e.g., GitHub) showing clear, well-documented code
• ML generalist who knows what scalable, reliable ML systems look like
• Experience with equivariant architectures, geometric deep learning, or GNNs (NequIP, MACE, SchNet, PaiNN, or similar), and/or ML-assisted drug discovery
• Experience building in declarative workflow orchestration frameworks like Flyte, Dagster, etc
• Lack of fear around interacting with quantum chemical scientists and their data pipelines

Qualifications

Must Haves

• At least two years relevant industry experience
• Highly fluent in and enthusiastic about PyTorch and JAX
• Used to thinking in asynchronous primitives
• Strong views on library design: clean abstractions, minimal surface area, consistency
• Solid track record of observable artifacts (e.g., GitHub) showing clear, well-documented code
• ML generalist who knows what scalable, reliable ML systems look like

Nice to Haves

• Experience with equivariant architectures, geometric deep learning, or GNNs (NequIP, MACE, SchNet, PaiNN, or similar), and/or ML-assisted drug discovery
• Experience building in declarative workflow orchestration frameworks like Flyte, Dagster, etc
• Lack of fear around interacting with quantum chemical scientists and their data pipelines