The posting
BigHat Biosciences is seeking an exceptional player-coach to serve as Director of Machine Learning. Reporting to the CIO, you will own the technical direction, and a meaningful share of the hands-on execution, of BigHat's machine learning capabilities and the growing domain of machine learning operations, including agent development and how agents automate and extend our internally built computational platform and force multiply the impact of our teams.
We are looking for someone with genuine theoretical depth and the vision to say where ML-driven antibody design should go next. Your team will develop state-of-the-art antibody property models with the goal of iteratively improving our antibody designs and you’ll own the proof points that demonstrate the impact of these models. You’ll build agent services to bolster the agility and impact of our teams and work with frontier AI labs to adapt LLMs to new agentic antibody design challenges. As our platform value grows, so does our opportunity to create new and valuable platform partnerships. You will be accountable for the ML that differentiates our therapeutic programs, and for the modeling capabilities we put in front of data and model partners.
At BigHat, we only declare success upon synthesis of real antibodies with drug-like properties that improve patient outcomes. This is a rare opportunity to run frontier ML within a one-of-a-kind production platform, with an automated high-throughput wet lab generating the data to fuel it.
Key Responsibilities
- ML strategy and vision: Own and continuously sharpen BigHat's machine learning roadmap. Articulate where the field is going and where we intend to lead.
- Deliver best in class therapeutic models: Utilize BigHat’s data repository and data generation capabilities to create multi-objective predictive models to demonstrate a durable, benchmarked, best-in-industry proof that better models accelerate real therapeutics. Develop and train models that support the safe and effective creation of increasingly sophisticated, multi-specific antibodies that patients need today.
- Frontier models joined with best-in-class data: Apply frontier models, utilizing BigHat’s unique datasets and data capabilities, against nature’s hardest protein design challenges. Optimize our design loop by integrating our lab, computational platform and data with LLM automation.
- Agentic direction: Join forces across the CIO organization to steer our unified LLM and ML approach. Relentlessly drive our adoption of safe, impactful LLMs and similar models, agent frameworks and RL evaluation benchmarking frameworks to empower every BigHatter.
- Program impact: Ensure that models, provenance, and benchmarks flow through our infrastructure without heroics, and that build-test-train design rounds fully leverage model insight for all 2k+ antibodies designed each week. Be accountable for computational deliverables to internal and partnered programs, delivering timely analyses, predictions, and optimized variant sets that drive progress toward clinic-ready molecules.
- Platform as product: Work with business development, partnerships, and executive leadership to translate our data generation and modeling capabilities into offerings that frontier AI companies and biopharma partners can buy. Scope pilots, define what technical success looks like, and deliver it.
- Team leadership: Mentor and grow a small, senior team of ML scientists and engineers. Set the bar for rigor, collaboration, and intellectual honesty, and create the conditions for people to do the best work of their careers.
- External leadership: Publish, present, and engage the ML and protein engineering communities. Strengthen BigHat's reputation as the place where frontier machine learning meets industrial-scale, purpose-built biological data.
Skills, Knowledge and Expertise
- Theory and vision: You reason from first principles about why a method should work, design the experiment that would falsify it, and explain the whole thing clearly to a room of interdisciplinary scientists. You hold opinions about where protein ML is heading and have a track record that supports them.
- Technical mastery: Deep, current command of modern ML: generative models of protein sequence and structure, Bayesian optimization and active learning, uncertainty quantification, data-efficient predictive modeling/in-context learning, etc. Strong Python and PyTorch, with real comfort in modern software engineering practice, cloud computing, and MLOps.
- Player-coach instinct: You lead by doing as well as by directing. You are energized by writing code and running benchmarks alongside your team.
- Operational excellence: You have shipped ML into production systems and care about reproducibility, provenance, benchmarking, and cycle time as much as about model architecture.
- Biological fluency: Working knowledge of antibody biology, developability, and drug development, or the demonstrated ability and appetite to acquire it quickly. Familiarity with display, NGS, and high-throughput biophysical data is a strong plus.
- Education & Experience: Ph.D. in Machine Learning, Computer Science, Computational Biology or a related field, with 8+ years of experience developing and applying advanced ML methods (industry experience strongly preferred), including demonstrated leadership of technical teams or major technical programs.
- Collaboration & Communication: Excellent written and verbal communication. You thrive in interdisciplinary settings alongside protein scientists, automation engineers, software engineers, and drug developers, and you are credible in front of partners and external audiences.
- Passion for Impact: Motivated by BigHat's mission to bring safer, more effective antibody therapies to patients, and motivated by the chance to make a one-of-a-kind platform the best in the world at what it does.
Total Rewards
The salary estimated for this position is $250,000 - $365,000 + bonus + options + benefits. Compensation will vary depending on job-related knowledge, skills, and experience. Actual compensation will be confirmed in writing at the time of the offer.
