Engineering Manager, Perception

Overview

We’re Blue River, a team of innovators driven to create intelligent machinery that solves monumental problems for our customers. We empower our customers – farmers, construction crews, and foresters - to implement safer and more sustainable solutions, driving increased profitability with less reliance on scarce labor. We believe that focusing on the small stuff – pixel-by-pixel and task-by-task - leads to big gains.

Blue River Technology aligns with John Deere’s vision to “innovate on behalf of humanity” by quickly identifying and solving high-value, high-uncertainty challenges in AI, machine learning, computer vision, and robotics. BRT acts as a research and development flywheel, building not only new products but also new platforms that reliably create value for both Deere and its customers. From fully autonomous machines to highly precise farming equipment, BRT and Deere are partnering to create technical breakthroughs in industries like agriculture and construction.

Our people are at the heart of what we do. Through cross-disciplinary collaboration, this mission-driven team is eager to define the new frontier of robotics. We are always asking hard questions, rapidly iterating, and getting our boots in the field to figure it out. We won’t give up until we’ve made a tangible and positive impact on the planet!

Blue River Technology is based in Santa Clara, CA.

We are developing autonomy for off-road construction vehicles, including autonomous articulated dump trucks (ADTs). This role owns the entire data, perception ML, and analytics stack, from raw sensor ingest to optimized edge inference and fleet-level performance visibility.

You will lead a small Data and CV/ML team while remaining deeply hands-on. You are accountable for how data is collected, curated, trained, deployed, measured, and improved in production. This is a founder-mode role with direct impact on autonomy performance in the field.

• Employment Type
  :
  Full-Time
• Visa sponsorship is considered for this position.

A combination, not necessarily all-inclusive, of the following:

• End-to-End Technical Ownership
  • Own the full perception ML stack, from data ingest and labeling strategy through training, validation, deployment, and on-vehicle inference.
  • Lead development of segmentation and depth estimation systems for off-road autonomy.
  • Design and ship production-grade models using Python and PyTorch.
  • Deploy and optimize models on NVIDIA Jetson / Orin, including Tensor
    
    RT optimization and runtime profiling.
  • Ensure models meet real-time, reliability, and resource constraints required for safety-critical systems.
• Own the data lifecycle: collection, mining, curation, labeling, versioning, and governance.
• Drive active learning pipelines to surface failure cases, edge conditions, and high-impact training data.
• Ensure dataset coverage aligns with operational risk, environmental diversity, and product priorities.
• Fleet Analytics & Performance Metrics
  • Own the definition, interpretation, and usage of autonomy performance metrics across the fleet.
  • Leverage and evolve the existing DFA Dune fleet analytics dashboard to ensure it reflects real-world robot behavior.
  • Partner with the data engineer responsible for analytics to:
    • Ensure metrics are accurate, actionable, and trusted
    • Tie perception and ML performance directly to in-field autonomy outcomes.
  • Use analytics to identify regressions, guide data collection, prioritize model improvements, and inform leadership.
• Manage a small Data team (1–2 data engineers/scientists) and CV/ML team (~3 engineers).
• Set technical direction, review designs and code, and maintain a high execution bar.
• Ensure the team understands how their work impacts real-world vehicle performance, not just offline metrics.
• Cross-Functional Communication
  • Clearly communicate technical progress, risks, and tradeoffs to leadership and partner teams.
  • Maintain a high signal-to-noise ratio in communication, especially around uncertainty and failure modes.
  • Act as the point of contact for perception and data on autonomy, systems, and field operations.
• Validation & Field Performance
  • Define evaluation and validation strategies…