דרושים Engineering Technician Infotainment Vehicle Software ב-Tesla ב-United States, Palo Alto

• Understand and model the end-to-end training performance of the Autopilot SW team's Pytorch-derived neural networks on the Dojo system
• Develop software that scales and improves training performance based on your analysis of the bottlenecks
• Collaborate with the Dojo HW team to understand current HW architecture and propose future improvements

• Degree in Engineering, Computer Science, or equivalent in experience and evidence of exceptional ability
• Experience scaling neural network training systems or other large distributed systems
• Familiarity with the internals of PyTorch and/or JAX
• Performance analysis experience
• Experience coding parallel programs
• Able to work from Palo Alto office

• Develop and maintain hardware libraries and drivers
• Assist test engineers in integrating applications into the framework
• Write drivers from scratch when needed
• Support the software framework and infrastructure
• Work on deployments and automation for various projects
• Collaborate with design validation and characterization teams
• Ensure seamless hardware-software integration
• Troubleshoot and resolve software and hardware-related issues
• Contribute to improving development workflows and tooling

• Bachelor's Degree in Computer Science, Electrical Engineering, Computer Engineering, or equivalent experience
• Strong experience with a statically typed language such as Go, Java, C, C++, or Rust
• Proficiency in developing hardware drivers and libraries
• Understanding of software-hardware interactions and integration
• Experience with wrapping low-level code into services
• Familiarity with deployment automation and CI/CD workflows
• Strong understanding of computer networking concepts, including VLANs and distributed system networking
• Ability to troubleshoot hardware and software integration issues
• Experience supporting framework or infrastructure development
• Strong problem-solving and debugging skills

Tesla AI’s Teleoperation team is charged with providing remote access to our robotaxis and humanoid robots. Our cars and robots operate autonomously in challenging environments. As we iterate on the AI that powers them, we need the ability to access and control them remotely. This requires building highly optimized low latency reliable data streaming over unreliable transports in the real world. At Tesla, we control the entire hardware and software stack, end to end. Our goal is to integrate our hardware, firmware and backend expertise to achieve a cutting-edge system. Our remote operators are transported into the device’s world using a state-of-the-art VR rig that allows them to remotely perform complex and intricate tasks. Working with hardware teams, you will drive requirements, make design decisions and implement software integration for this custom teleoperation system.

• Responsible for developing the Unreal application that our Remote Operators use to interface with our cars and robots
• Influence architectural and UX decisions with a focus on usability, security, reliability, and high performance
• Choose technologies that will allow us to scale with our growing fleet of cars and robots while efficiently utilizing cloud infrastructure and managing costs
• Setup and maintain monitoring, metrics & reporting systems for fine-grained observability and actionable alerting

• Practical experience using application layer languages like C++ and Python
• Expertise with 3D graphics and game engines – specifically Unreal
• Experience with both Windows and Linux environments: networking, wireless (WiFi and Bluetooth), filesystems and storage
• Familiarity with writing and maintaining custom shaders
• Familiarity with build systems like MSVC and bazel

• Low-level interface with camera image sensors and raw data analysis

• Development of frame grabbing interfaces and low-latency processing on embedded devices

• Debugging operating mode and electrical issues with cameras such as synchronization and interference issues

• Design and implement backend services and tools that handle fleet data collection, batch processing, training, simulation, and evaluation based on real-world sensor data

• Work closely with the other Autopilot teams to integrate with other tools and provide the most efficient software development process

• Degree in Computer Science, Electrical Engineering, or related area with strong software engineering experience, or equivalent experience

• Experience with C/C++ and Python

• Experience working in a Linux command line environment

• Experience with Docker

• Experience with container orchestrators and cluster management software preferred

• Experience with data stores and indexers like PostgreSQL, Elastic Search, Redis preferred

• Experience with Go is preferred

• Experience with data-visualization platform like Garfana is preferred

• Rapidly prototype in-car customer experiences

• Extract meaningful data from fleet telemetry via automated processes

• Implement complete features with both client and server footprint
  

• Drive your changes end to end from concept to in car drive test

• Design, code, and debug software on Linux-based embedded platforms and Linux servers

• Work with a cross-functional team of hardware engineers, application/UI software engineers, QA/Validation, and designers

• Degree in Computer Science, Information Systems, Electrical Engineering or the equivalent in experience

• Evidence of exceptional achievement in large-scale algorithmic environments

• Proficient in one of C++, Python, Golang, Scala

• Experience writing and shipping customer-facing code

• High standards for code quality, maintainability, and performance

• Willingness to learn new technologies in any area of the stack: Vehicle Control, Routing/Guidance Algorithms, Server Infrastructure, UI, Embedded Systems

• Experience with Maps / GIS / Computational Geometry data is a strong plus

• Experience in Spark, AWS, Kubernetes, Jenkins is a plus

• Experience in Machine Learning algorithms is a plus
  

• Design and implement backend services and tooling that handles iteration and batch processing of inference, simulation, and evaluation workloads
• Work closely with the other Autonomy teams to build foundational components as well as bridge missing pieces in ML compiler and runtime infrastructure while designing for scalability, reliability, security and high performance

• Proficiency with Python
• Familiarity of managing hardware inference chips like TPUs and of optimizing machine learning inference workloads for low latency and scale
• Familiarity with Operating Systems concepts such as networking, processes, file systems and virtualization
• Familiarity with concurrency programming
• Familiarity with C++ and / or Golang
• Experience with Linux, container orchestrator like Kubernetes or similar and bare metal setup tools like Ansible or similar
• Experience with data stores like PostgreSQL and Redis

• Ship reliable, cutting-edge generative AI applications on Tesla's in-house platform
• Improve & maintain the underlying components of the generative AI platform
• Extract as much performance and utility out of specialized hardware as possible
• Help solve a wide variety of far-reaching problems using modern AI- and ML-based techniques
• Work in the challenging space of transferring concepts between research ideas to productionized services
• Deliver products and services that have wide-ranging impacts for their users

• Experience delivering real-world generative AI applications, especially ones based around Large Language Models (LLMs)
• Practical experience with the Python ML ecosystem (Tensorflow, PyTorch, JAX, LangChain, TGI/vLLM, etc.)
• Knowledge of Python internals for techniques necessary to scale to large numbers of concurrent users
• Proficiency in taking ideas from research prototypes and deploying them into productionized services
• A track record of shipping reliable applications that require minimal support and incur minimal downtime
• A healthy respect for inherent complexity and a deep suspicion of accidental complexity