Senior Machine Learning Infrastructure Engineer
Roblox · San Mateo, CA, United States
About this role
**Senior Machine Learning Infrastructure Engineer (Roblox)** Roblox is building the tools and platform that empower its community to bring any experience to life—at massive scale. Matchmaking is a critical stage downstream of Discovery and Search: once a player chooses a game, matchmaking assigns them to a server, shaping who they play with, the latency they experience, and the quality of their session. As a **Senior Machine Learning Infrastructure Engineer**, you’ll build scalable, reliable, high-performance infrastructure that powers ML systems across the organization—at the scale of **tens of billions of scorings**, and helping match **hundreds of millions of users** to **millions of games**. --- ### You will - Work at the intersection of **model architecture** and **ML infrastructure**, partnering with core modelers, data/AI infrastructure engineers, and product teams to push large-scale training and serving. - Design and own large-scale **data engineering and streaming systems** for ML, including real-time event ingestion, streaming feature computation, training-data generation, and data quality/freshness guarantees. - Build low-latency pipelines using **Kafka, Flink, Spark Structured Streaming, Ray**, and keep online/offline features consistent. - Build **online and continual learning** systems so models adapt to user behavior and content shifts in near real time. - Include incremental/streaming training, online updates/fast refresh, embedding + parameter-server synchronization, delayed-feedback/label-joining pipelines, and safe deployment with monitoring, drift detection, and automated rollback. - Investigate end-to-end **model, data, and systems tradeoffs** from streaming pipelines and distributed training to low-latency inference and production serving. - Design cache strategies; apply pruning/quantization; optimize GPU utilization and memory efficiency; develop custom kernels when needed. - Own ML infrastructure architecture and establish best practices for **scalability, reliability, and cost-effectiveness** (training, serving, features). - Bring production experience with low-latency key-value stores and feature registries such as **Redis, Feast, Tecton, Cassandra**. - Partner with data scientists, ML engineers, platform teams, and product stakeholders to design, implement, and operate robust ML platforms. - Stay current on industry trends to adopt leading-edge ML infrastructure technologies and practices. --- ### You have - **5+ years** experience designing, building, and deploying large-scale ML and data systems in production. - **3+ years** experience tech-leading ML infrastructure or data engineering teams. - Hands-on experience with large-scale streaming/data processing (e.g., **Kafka, Flink, Spark, Beam**) and real-time feature pipelines. - Experience building/operating online, incremental, or continual learning systems for recommendation, ranking, or search is a plus. - Strong communication skills and a collaborative approach. - **BS/MS/Ph.D.** in Computer Science, Engineering, or equivalent experience. --- ### Pay & Location - **San Mateo, CA (HQ):** Annual base salary range **$243,290 — $295,250 USD** (actual base pay depends on factors such as background, experience, location, and market demand). Equity and benefits are also available. - Office schedule: onsite **Tuesday, Wednesday, Thursday** (optional presence Monday/Friday unless otherwise noted). --- Roblox is an equal opportunity employer and
Listing freshness
CronJobs last confirmed this listing 57m ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.