Principal Machine Learning Engineer, Matching and Recommendations
Ai Ml
Artificial Intelligence
Big Data
Data Analysis
Data Architecture
Data Integration
Data Pipeline
Data Processing
Data Science
Engineer
Generative AI
Large Language Models
Machine Learning
Machine Learning Engineering
Machine Learning Evaluation
Machine Learning Experimentation
Machine Learning Modeling
Machine Learning Pipelines
Programming
Recommendation & Personalization
Recommendation Engines
Recommendation Systems
Job Description
Bumble is seeking a Principal Machine Learning Engineer to define and lead ML strategy for next-generation matching and recommendation systems.
Responsibilities
- Define and lead technical strategy for AI and ML systems that support recommendations, ranking, and personalization across Bumble products
- Deliver measurable improvements in user engagement and safety
- Design, develop, and deploy production-grade models using modern ML frameworks such as PyTorch, with focus on scalability and reliability in high-traffic environments
- Build and deploy production AI Agents using raw and fine-tuned foundational LLMs, including sub-agents, tools, and MCP integrations
- Architect end-to-end ML pipelines, integrating data processing (e.g., Spark, Airflow) with training, evaluation, and deployment workflows
- Drive experimentation frameworks, including A/B testing and offline evaluation, to improve model performance and product outcomes
- Partner cross-functionally with Product, Engineering, and Data leadership to convert business challenges into impactful ML solutions
- Mentor and elevate senior individual contributors, supporting a culture of Excellence, Curiosity, and continuous learning across the ML community
- Own complex and ambiguous problem spaces from insight to impact, adapting with an agile mindset
- Champion responsible AI practices to embed fairness, transparency, and user safety into ML systems
Requirements
- Typically requires 10–15 years of experience (alternative backgrounds considered when skills are equivalent)
- Deep, hands-on expertise building and deploying large-scale ML systems in production
- Python proficiency and experience with at least one major ML framework (e.g., PyTorch or TensorFlow)
- Experience in recommendation systems, ranking models, and/or NLP
- Expertise in prompting and fine-tuning LLMs, plus building production AI Agents
- Proven ability to design scalable data and ML pipelines using tools such as Spark, Airflow, or similar distributed systems
- Ability to operate as a senior individual contributor and influence technical strategy and decisions without direct authority
- Demonstrated cross-functional collaboration and ownership of outcomes in complex organizational environments
- Track record mentoring and uplifting others, role modeling Respect and Excellence while building inclusive, high-performing teams
- Strong AI fluency to independently design, evaluate, and optimize ML systems and guide responsible, effective AI use by others
Technologies
- Python
- PyTorch
- TensorFlow
- Large Language Models (LLMs)
- MCP
- Spark
- Airflow
- A/B testing
Compensation
- USD 345,000 - 410,000 per year
- Final compensation determined based on qualifications, relevant experience, skill set, and other job-related factors
Inclusion at Bumble Inc.
- Bumble Inc. is an equal opportunity employer and encourages candidates of all ages, colors, sexual orientations, gender identities, veterans, parents, people with disabilities, and neurodivergent people to apply
- Reasonable adjustments are available to help candidates feel more confident during the process
- Candidates may include pronouns in the application (e.g., she/her, he/him, they/them)
AI in Bumble Inc. Hiring
- AI tools may support parts of recruitment, such as recording, transcribing, summarizing conversations, and highlighting resume-job alignment
- Hiring decisions are made by people; AI supports efficiency and candidate experience, not evaluation or decision-making
- AI-supported interviews and conversations are voluntary and do not affect candidacy
- Opt out is available by informing the recruiter or interviewer at the start of a call or anytime during the conversation
- Summaries and related data are retained only as needed per internal retention policies; transcription or summary deletion can be requested via the recruiter
Fraudulent Candidate Detection
- The applicant tracking system analyzes signals tied to device, IP, email, and phone data to protect applicants and the hiring process
- Signals are for internal review only and are not a definitive determination of identity or intent
- No rejection, advancement, or impact occurs based solely on this analysis without human review
Location: New York, NY (onsite)