About the Role
Design and deploy machine learning solutions that improve the accuracy and coverage of product recommendations by optimizing recall in large-scale search systems.
Responsibilities
- Develop high-performance retrieval models to increase coverage of relevant results
- Optimize embedding-based search architectures for low-latency inference
- Collaborate with data scientists to refine training datasets and evaluation metrics
- Improve model freshness by streamlining retraining pipelines
- Work closely with backend engineers to integrate ML components into production systems
- Diagnose gaps in recall and implement targeted model improvements
- Evaluate trade-offs between precision, recall, and computational cost
- Monitor model performance in production and respond to degradation
- Contribute to the development of evaluation frameworks for retrieval systems
- Implement efficient indexing strategies for large-scale vector databases
- Design experiments to measure the impact of recall improvements on business metrics
- Stay current with advancements in deep learning and information retrieval
- Document technical designs and share knowledge across the ML team
- Support A/B testing initiatives involving retrieval algorithms
- Ensure models comply with data privacy and security standards
- Refactor legacy ML code for better maintainability and scalability
- Assist in defining long-term roadmap for recall-related ML capabilities
- Troubleshoot data flow issues across distributed systems
- Optimize resource usage of ML services in cloud environments
- Mentor junior engineers in best practices for machine learning engineering
Nice to Have
- Master’s or PhD in computer science or related discipline
- Experience with e-commerce or product search domains
- Contributions to open-source ML or search projects
- Publications or presentations in ML or IR conferences
- Deep expertise in semantic search or cross-modal retrieval
Compensation
Competitive salary and equity package
Work Arrangement
Remote-friendly with team collaboration expectations
Team
Part of a specialized machine learning team focused on search relevance and personalization
About the Role
This position focuses on advancing the recall component of a product search engine using machine learning. The engineer will work on improving the system's ability to surface relevant items from a large catalog by refining retrieval models and data pipelines. Success in this role means delivering more comprehensive and accurate search results through scalable ML infrastructure.
Impact
The work directly influences the quality of search results for millions of users. Improvements in recall lead to better product discovery, higher customer satisfaction, and increased conversion rates. Engineers in this role have the opportunity to see their models impact real-world user behavior quickly.
Technology Stack
The team uses Python for model development, PyTorch for deep learning, and Apache Airflow for workflow orchestration. Vector search is powered by approximate nearest neighbor libraries. Infrastructure runs on Kubernetes with cloud-based storage and compute.
Collaboration
Engineers work closely with data scientists, backend developers, and product managers. Regular syncs ensure alignment on goals, metrics, and deployment timelines. Cross-team knowledge sharing is encouraged through tech talks and documentation.
Growth Opportunities
The role offers paths for technical leadership, including ownership of core ML subsystems and mentoring junior team members. Engineers are supported in attending conferences and pursuing advanced training.
Work Environment
The team values autonomy, technical rigor, and continuous learning. Engineers are expected to take initiative while collaborating effectively. Remote work is supported with asynchronous communication norms.
Performance Metrics
Success is measured by improvements in recall-at-k, reduction in false negatives, and positive shifts in user engagement metrics. Model stability and latency are also monitored in production.
Onboarding
New hires receive a structured onboarding plan including system walkthroughs, access setup, and initial project assignments. A mentor is assigned to guide the transition into the team.
Available for qualified candidates
