Senior Machine Learning Engineer

Role Overview

Senior Machine Learning Engineer to lead the architectural evolution of Samsara’s safety systems. You will move from siloed end-to-end models to a unified Perception Platform Layer, building infrastructure that converts raw sensor data into real-time, safety-critical decisions across cloud and edge.

Responsibilities

• Platform Architecture & Unification: Lead the transition from fragmented, task-specific models to a modular perception platform with reusable components and downstream safety applications.
• System Design: Design and implement real-time ML systems, covering the pipeline from sensor ingestion and tracking through risk reasoning and actuation, with clear interfaces and predictable behavior.
• Hybrid Deployment: Integrate and orchestrate models across edge and cloud, including versioning, rollouts, and mission-critical fallback mechanisms.
• Performance & Reliability Engineering: Own end-to-end latency and reliability for safety-critical pipelines; profile, schedule, and optimize messaging and backpressure across the stack.
• Observability & Feedback Loops: Build monitoring to detect model drift, false positives/negatives, and latency regressions; close the loop so production data informs training.
• Rigorous Evaluation & Safety: Develop evaluation frameworks for rare long-tail safety scenarios (text cuts off mid-sentence).

Requirements / What You Should Apply If

• Strong motivation to create real-world impact in industries powered by physical operations (e.g., safety outcomes).
• Ability and desire to build for scale (Samsara has millions of deployed IoT devices).
• Growth mindset and willingness to learn across technologies, challenges, and customers.
• Customer empathy and comfort collaborating with customer success and product managers; use metrics to improve outcomes.
• Team-oriented collaboration while executing independently.

Nice-to-haves

• Experience designing safety-critical or mission-critical real-time ML systems with hybrid edge/cloud deployments.
• Experience with monitoring/observability, drift detection, and production-to-training feedback loops.

Location / Work Model

• Remote (US-based candidates).