Build science
from first principles.
We're assembling a small, high-impact team of scientists and engineers who want to tackle one of medicine's hardest problems. Early-stage means ownership, speed, and working on things that matter.
What We Value
Deep Curiosity
We ask hard questions, challenge assumptions, and go deep on problems before reaching for solutions. The best science starts with the right questions.
Cross-Disciplinary
The intersection of ML and wet lab chemistry is where our edge lives. We value people who can move fluidly between computation and experiment.
Bias for Impact
We move fast, iterate relentlessly, and stay focused on what matters: building a platform that changes how drugs reach patients.
Current opportunities.
AI Scientist
Computation
Philadelphia · Hybrid
+
Develop novel algorithms and machine learning architectures that accelerate the drug discovery and delivery process. A high-impact, cross-functional role with close collaboration between lab scientists, software engineers, and business leaders. You'll design and build foundational models, work directly with wet lab teams, and translate biological challenges into computational solutions.
Key Responsibilities
- Develop foundational generative models for designing novel chemical structures and compounds for drug and vaccine delivery systems
- Design, build, and scale production-grade machine learning models
- Contribute to building and growing the computational and machine learning team
- Collaborate closely with the wet lab and platform teams to improve and iterate on therapeutic applications
- Establish and maintain best practices for model development, validation, and deployment
- Communicate findings and progress to technical and non-technical stakeholders, including leadership
What We're Looking For
- Master's or PhD in Computer Science, Machine Learning, or a related field
- 2+ years of experience building and deploying production machine learning models
- Advanced proficiency in Python
- Experience with ML frameworks like PyTorch, TensorFlow, or JAX
- Strong mathematical foundations and deep understanding of ML principles, network architectures, evaluation metrics, and optimization
- Solid background in software engineering and algorithm design
- Ability to thrive in fast-paced, dynamic environments with evolving objectives
- Interest in computational chemistry, biology, and healthcare
Nice to Have
- Experience applying ML to biological or chemical domains
- Familiarity with state-of-the-art generative models (LLMs, diffusion & flow models, VAEs, GANs, EBMs)
- Publications at top-tier conferences such as NeurIPS, ICML, ICLR
- Familiarity with wet lab processes and experimental pipelines
Senior Scientist, Protein Design
Computation
Philadelphia · Hybrid
+
A PhD-level scientist with deep expertise in computational and ML-driven protein binder design. You'll thrive at the interface of sequence/structure-based protein design and experimental validation, with work that directly influences therapeutic discovery programs. You'll develop and apply advanced computational approaches to design and optimize protein binders, working closely with experimental scientists to translate models into testable molecules.
Key Responsibilities
- Lead independent computational research focused on protein and binder design for therapeutic applications
- Develop, implement, and refine sequence- and structure-based models to guide binder discovery and optimization
- Deliver robust, reusable computational methods and workflows that support iterative experimental testing
- Collaborate closely with experimental scientists to ensure designs are aligned with biological constraints and assay capabilities
- Interpret experimental results and incorporate feedback into subsequent design cycles
- Communicate scientific insights, limitations, and progress clearly to cross-functional teams and leadership
- Contribute to building internal best practices for computational protein design, data management, and model evaluation
What We're Looking For
- PhD in computational biology, computational chemistry, machine learning, computer science, or a closely related field
- Demonstrated experience developing novel AI methods, including sequence- and/or structure-based approaches for biomolecular design
- Strong foundation in protein structure, molecular interactions, stability, and biophysical constraints
- Hands-on experience with modern protein design and modeling tools (AlphaFold, ProteinMPNN, Rosetta-based workflows, diffusion or generative models for proteins)
- Proficiency in Python and modern ML frameworks (PyTorch, TensorFlow, or JAX)
- Ability to work independently and contribute effectively within a collaborative, interdisciplinary team
Nice to Have
- Prior industry experience applying computational methods to therapeutic protein or binder discovery
- Familiarity with integrating computational outputs into wet-lab workflows
- Strong scientific communication skills
- Track record of publications in protein engineering, structural biology, or computational biology
Don't see your role?
If you're exceptional and believe in what we're building, we want to hear from you — regardless of whether a formal role is listed.
contact@koshasciences.com