Software Engineer - ML/DevOps

Toronto, Ontario /
Engineering – Machine Learning /
Full-time
BenchSci's vision is to help scientists bring novel medicine to patients 50% faster by 2025. We do this by empowering scientists to run more successful experiments with the world's most advanced, easy to use biomedical artificial intelligence software platform, thereby avoiding delays that slow the progress of medicine to clinical trials. Backed by F-Prime, Inovia, Golden Ventures, and Google's AI fund, Gradient Ventures, we provide an indispensable tool for more than 41,000 scientists that accelerates research at 15 top 20 pharmaceutical companies and over 4,300 leading academic centers. We're a CIX Top 10 Growth company, certified Great Place to Work®, and top-ranked company on Glassdoor.

We are seeking a Software Engineer - ML/DevOps to join our growing R&D team! Reporting to the Engineering Manager, R&D, you will operationalize machine learning experiments (notebooks, scripts) into production-grade deployments (tracked and monitored ML pipelines). Successful candidates for this role are well-versed with the nuances of CI/CD in the machine learning lifecycle along the dimensions of code, data and model reproducibility, and possess a results-oriented approach to measure the reduction of complexity to their peers’ engineering workflows. Ultimately, the mission of this role is to minimize the time it takes to bring a model to production.

You Will:

    • Drive core components of our ML technical roadmap concerning high-throughput online inference, active learning for supervised approaches, and scalable & reproducible training, to name a few.
    • Create and implement reference architectures for both training and inference at scale, collaborating and knowledge-sharing with our Core Infrastructure team to compress time-to-production for Machine Learning
    • Build tooling and pipelining abstractions to allow ML engineers to focus on experimentation while empowering self-service to deploy and serve their models reliably and consistently
    • Help ML engineers produce clean, tractable, and highly performant machine learning through rigorous code review with a lens on software quality
    • Advocate for code and process improvements across the Machine Learning team, and help to define best practices based on personal industry experience and research
    • Participate in sprint planning, estimations, and reviews

You Have:

    • 3+ years of software engineering experience, preferably in Python
    • Experience with maintaining functional, production reference architectures for end-to-end Machine Learning in cloud, on-premise, and hybrid environments
    • Strong Linux system administration skills 
    • Familiarity with industry standards such as MLFlow and DVC
    • Experience with declarative infrastructure and Kubernetes (GKE) for model serving and scalable inference
    • Exposure to automated testing and CI/CD in the ML context
    • An understanding of fundamental ML concepts like metrics, biases, and datasets
    • A constant desire to grow and learn
    • Strong cross-team communication and collaboration skills
    • A desire to see teammates succeed together 

Nice to haves, but not mandatory qualifications:

    • Experience with TensorFlow, PyTorch, and image processing libraries such as OpenCV and scikit-image
    • Experience with Google AI Platform for PyTorch implementations
    • Background in Life Science
    • Experience with Spark 2.x or later
    • Experience with pipelining/orchestration frameworks such as Airflow or Beam
    • Experience with hermetic build tools such as Bazel

Our benefits and perks:

    • A compensation package that includes equity options in the company
    • An annual Executive Health Assessment at Medcan: All employees get the “executive treatment”
    • Effectiveness coaching for managers: Onsite, personalized coaching from an executive coach with a doctorate in clinical psychology
    • Mental health tools and support: Optional mindfulness sessions and a free Headspace account
    • Complimentary genome sequencing from 23andMe: Find out what your DNA says about your health, traits, and ancestry
    • Three weeks of vacation, plus another week: Get 15 days to use anytime, and we’re closed Dec 25-Jan 1
    • Additional days off: Company summer day, your birthday, and earn +1 vacation day annually
    • Work from anywhere flexibility: Every day right now, and up to 4 days per week once we return to the office
    • An onsite gym: Keep fit, conveniently, with a Peloton and other great equipment
    • A great benefits package: Including health and dental
Here at BenchSci, these are our core values:

Focused: We focus on what will drive the greatest impact at all times.
Advancement: We believe in continuous growth, and discovering new ways to do things better. This applies to our product and business, but also to ourselves.
Speed: We recognize that without a sense of urgency, our team, our product and our mission lose their value.
Tenacity: What we’re trying to do isn’t easy, but we hire the best people, and give them the autonomy, tools, and resources to succeed. The hard work is up to them.
Transparency: We believe that sharing diverse ideas and information creates strong teams. Our success stems from research, collaboration, feedback, and trust.

Diversity, Equity and Inclusion:
BenchSci is committed to creating an inclusive environment where people from all backgrounds can thrive. The work and commitment to diversity, equity and inclusion is our collective responsibility. That fundamental belief will guide us along our diversity, equity, and inclusion journey. We are just at the beginning, we will experience moments of discomfort and we may stumble along the way but we are committed to continuously improving and creating equitable and systemic change.

Accessibility Accommodations:
BenchSci provides accessibility accommodations during the recruitment process. Should you require any accommodation, we will work with you to meet your needs.