Описание
Responsibilities
- Own, debug, and improve production pipelines for accuracy and efficiency.
- Design, build, and iterate on LLM evals and LLM-as-a-judge systems, including the datasets, metrics, and benchmarks to monitor quality.
- Run rigorous experiments on prompts, pipeline design, model choice, and inputs, and turn the results into clear recommendations.
- Drive LLM efficiency improvements across our pipelines by comparing models, restructuring steps, and replacing LLM calls with smaller, purpose-built models where it makes sense.
- Expand our anonymization pipeline, including evaluating and adapting existing models.
- Work with Python and quality engineers to take experiments into production, and share responsibility for the code you ship.
- Document findings in writing (e.g., experiment reports, design docs).
Requirements
- Solid machine learning/data science background: you understand experimental design, data splits, metrics, bias, and statistical significance.
- Hands-on experience applying LLMs in production or near-production settings, especially evaluating them: building eval sets, LLM judges and benchmarks, and iterating on prompts and pipelines.
- Comfortablility getting deep into messy data: exploring it, troubleshooting issues, and transforming insights into measurable improvements.
- Strong Python and good engineering skills. You write clear, readable, tested code.
- Able to work independently on ambiguous problems: framing the question, choosing the approach, and deciding on the acceptance criteria to ship.
- Use AI coding tools effectively and stay accountable for the output.
- Communicate clearly in spoken and written English.
Nice to have
- Experience with classical ML or deep learning model development.
- Experience with image/OCR pipelines or face-related anonymization.
- Experience shipping end-to-end ML features in a startup or scale-up environment.
We offer
- Generous compensation + stock options - aligned with our internal framework, market data, and individual skills.
- Distributed work: Work from anywhere - fully remote, in our hubs, or a mix.
- Company-issued laptop, remote setup stipend, and co-working budget
- Flexible schedules and location
- Ample paid time off, in addition to local public holidays
- Enhanced parental leave
- Health & retirement benefits
- Annual learning & development budget
- Annual workaways and regular virtual & in-person socials
- Opportunity to contribute to groundbreaking projects that shape the future of work
Hiring process
Apply