AI/ML Engineer, Peptide Properties and Binding ML

Amide Technologies•Waltham, MA

13d•Onsite

About The Position

Amide Technologies is a Massachusetts-based biotech company that designs therapeutic peptides and proteins which are difficult to obtain by conventional means, combining solid-phase peptide synthesis with biological expression. By using non-natural amino acids and artificial protein backbones, we design peptides that conventional chemistry and biology cannot reach. Because Amide makes compounds that have never been made before, the work regularly involves problems with no established playbook. We are seeking a technically strong AI/ML engineer to develop models that improve how Amide designs peptides for binding, structure and developability-relevant properties. This role will focus on practical machine learning for peptide sequence, structure, target engagement and key property workflows such as permeability, tissue distribution or half-life where data support meaningful modeling. You will work in a collaborative matrix environment with experimental scientists, computational biologists, chemists and platform engineers at our Waltham, Massachusetts site. Role description: The qualified individual will have recognized expertise in machine learning for protein, peptide, molecular or structural data, combined with a strong understanding of biological validation. The candidate will build models that prioritize peptide designs, interpret binding and property signals, and support prospective design-build-test-learn cycles. This role goes beyond model training and requires strong scientific judgment about which predictions are credible, actionable and worth testing experimentally.

Requirements

AI/ML Engineer, Computational Scientist, or Principal Scientist with a PhD, or MSc with substantial industry experience, in Computational Chemistry, Biophysics, Computer Science, Computational Biology or a related discipline.
Strong programming skills in Python and hands-on experience with modern ML frameworks such as PyTorch, JAX, TensorFlow or related tools.
Deep expertise in protein, peptide, molecular or structural modeling, with direct experience in binding prediction or sequence-structure-function modeling.
Experience with biophysical modeling, structural bioinformatics, molecular simulation, geometric deep learning, protein language models or related approaches.
Demonstrated ability to evaluate model quality in a scientific setting, including validation strategy, uncertainty, bias and prospective performance.
Strong understanding of how experimental data quality, assay design and synthesis constraints affect ML model usefulness.
Experience working in matrixed teams of experimental and computational scientists to meet project objectives.
Clear communication style, strong organizational skills and the ability to explain modeling decisions to non-specialist scientific stakeholders.
Experience with peptide or protein sequence representations, embeddings, structure-derived features or featurization strategies for ML.

Nice To Haves

Experience with peptide therapeutics, constrained peptides, macrocycles, non-natural amino acids or synthetic peptide design.
Experience modeling peptide and miniprotein developability-relevant properties such as proteolytic stability, solubility, half-life or aggregation risk.
Familiarity with active learning, Bayesian optimization, uncertainty estimation or other methods for iterative design cycles.
Experience using public protein structure or interaction resources, such as AlphaFold, PDB, UniProt, ChEMBL or related datasets.
Ability to benchmark emerging protein foundation models and adapt them to sparse, proprietary peptide datasets.

Responsibilities

Design and train machine learning models for peptide structure, binding, sequence-function relationships and property prediction.
Build practical modeling workflows that help prioritize new peptide designs for synthesis and experimental validation.
Apply structural bioinformatics, biophysical modeling and deep learning methods to understand peptide-target interactions.
Evaluate model performance using rigorous prospective and retrospective validation.
Work with assay analytics and experimental teams to convert model predictions into testable design hypotheses and learning-loop readouts.
Help establish at least one property-modeling focus area, such as solubility, half-life or another program-relevant peptide property.
Produce reproducible modeling workflows, well-documented datasets and clear technical summaries for project and leadership decisions.