Remote opportunity at
PostEraMachine Learning Researcher - Agentic Science
PostEra describes itself as an AI-first biotech business. Its Proton platform, built for medicinal chemistry, is used to find new medicines more quickly. The company's internal…
Career Tools
About This Role
PostEra describes itself as an AI-first biotech business. Its Proton platform, built for medicinal chemistry, is used to find new medicines more quickly. The company's internal pipeline is aimed at women's health and fertility conditions, where Proton has helped nominate several clinical candidates. PostEra also runs small molecule projects with drug company partners and reports more than $1 billion in AI collaboration deals, four of them deals spanning several years…
Job Description
PostEra describes itself as an AI-first biotech business. Its Proton platform, built for medicinal chemistry, is used to find new medicines more quickly. The company's internal pipeline is aimed at women's health and fertility conditions, where Proton has helped nominate several clinical candidates. PostEra also runs small molecule projects with drug company partners and reports more than $1 billion in AI collaboration deals, four of them deals spanning several years with Pfizer and Amgen. A separate antiviral discovery center, preparing for future pandemics, is funded by a grant the posting describes as among the largest NIH awards on record.
This remote, full-time position is for a machine learning researcher who will develop the company's agentic research track. The work involves agent-driven systems that build mechanistic models of biochemical and physiological behavior, plus analysis of biological datasets to support target validation. Researchers partner with chemists and biologists so their models shape drug discovery choices. A major emphasis is in-context learning across molecular and tabular data, along with foundation models trained on PostEra's proprietary multimodal information.
Responsibilities span the entire research cycle: framing problems, assembling datasets and evaluation episodes, establishing baselines, scaling training, running careful ablations, and moving successful techniques into tools that scientists rely on. No background in drug discovery is needed, though a real interest in learning the field is expected. The posting describes a deliberately flat structure in which recognition and pay reflect individual contributions.
Responsibilities
- Build and test agent-driven systems that automate quantitative modeling of biological and physiological processes
- Examine biological data to strengthen target validation work
- Drive self-directed projects on in-context learning, learning from few examples, tabular foundation models, and molecular ML
- Build models able to adapt to fresh assays, endpoints, biological targets, or compound families with scant labeled data
- Measure novel methods against robust baselines using carefully designed test sets
- Link modeling goals and metrics to decisions on potency, ADME, selectivity, lead optimization, and early trial design
- Engineer efficient training pipelines and scale models across many molecular and tabular tasks
- Produce readable, reproducible code and take part in review and documentation
Requirements
- A PhD in machine learning, or in a STEM research field centered on new ML methods
- A record of strong research, evidenced by publications or open-source work
- Solid background in modern ML, deep learning, or statistical modeling, including the theory behind the methods
- Demonstrated depth in at least one of: agentic science workflows, ML for bioinformatics or clinical data, in-context learning, tabular learning, or few-shot learning
- Practical experience training, debugging, and assessing ML models in Python with PyTorch or JAX
- Skill in turning unclear scientific questions into concrete ML projects with datasets, baselines, metrics, and validation plans
- Comfort with startup conditions where goals change and data is messy
Qualifications
- Experience training tabular foundation models, particularly on sparse, high-missingness datasets
- Hands-on work on in-context learning for molecules, or adapting general-purpose models to scientific datasets
- Track record training 1B+ parameter networks across multiple machines with data or model sharding
- Prior work on co-scientist style AI tools for physical or life science problems
- Experience applying ML to biological, biochemical, or clinical data
- Moving research prototypes into production scientific tools
Core Skills
Benefits
- Equity of 0.05% to 0.1%
Frequently Asked Questions
Answers are based only on the employer’s listing; where it doesn’t say, neither do we.
Is this role remote?
The posting lists the role as remote and full-time.
What salary does the posting state?
The posting states a salary of 200k to 300k, plus equity of 0.05% to 0.1%.
Is visa sponsorship offered?
The posting says visa sponsorship is not available at this time.
What degree is required?
The posting requires a PhD in machine learning or a STEM research field involving new machine learning methods.
Sample Interview Questions
AI-generated questions tailored to this specific role — a preview of the full practice set.