Overview
Skills
Job Details
Role: Machine Learning Engineer
Location: remote - Eastern Time
Contract Fulltime with the client
Team Structure
Teams: Data engineers, data scientists, MLOps engineers (all split up)
New Role: With CVML team (2 engineers, both US employees reporting to Tarik in the UK)
Focus: Incremental improvements to existing models, continuous training operations (different from research engineers)
Tasks: Data exploration from the field, model development, deploying models
Technical Requirements
Languages: Python and C++
Experience:
Edge ML hardware and training models on the cloud (6 months on edge ML, 2+ years in the industry)
Relevant modules or experience during university degree
Skills:
Able to be mentored but not to mentor others
Not looking for genAI/NLP experience
Project Details
Focus: Design next set of intrusion detection series and capture thefts (hardware and software)
Validate chipsets and design chips for inference
Train new models for intrusion detection
Patent new solutions, sensors between camera and radar
Create privacy-compliant computer vision systems (e.g., face blurring, anonymization)
ML Models:
Convolutional Neural Networks (CNNs) are primary (YOLO series, MobileNet)
Frameworks: TensorFlow, PyTorch, ONNX (PyTorch can be learned on the job)
Experience with other image signal processing techniques
Less interest in LLMs/transformer experiences
Data Challenges:
Identify useful data among redundant data
Use ML to determine valuable data
Address edge cases (e.g., theft detection in dark conditions)
Domain Knowledge:
Can be learned on the job (focus on CV, image processing)
Model Deployment:
Shrink and compress models (quantization is key)
Familiarity with neural architecture search is a plus but not required
Scripting:
Basic Unix or shell scripting (not a major part of the role)
Signal Processing:
Generalist skills over time (computer vision to radar and vice versa)
Experience with filtering and time series data techniques
Tools:
ONNX for model conversion
AWS (EC2, S3, SageMaker); Google Cloud Platform is acceptable
ML operations tools (MLflow, Weights & Biases)
Candidate Screening Process
Steps:
Phone call (30 minutes)
Technical assessment (1.5 hours)
Behavioral interview (1 hour)
Responsibilities:
Intro call and technical piece handled by the recruiter
Manager will provide technical questions
Assessment:
Coding test to evaluate skillset
Data science question involving SQL and data manipulation in Python
Open-ended problems to assess problem-solving and thought process
Stock JD
Description:
As a Machine Learning Engineer in the AI squad, reporting to the Director of AI Engineering, you ll focus on developing AI-driven solutions to combat vehicle and content theft. This role is critical for driving innovation in our security technologies, helping to safeguard assets and push the boundaries of what s possible in the field. If you re passionate about solving real-world challenges with cutting-edge solutions, this is the opportunity for you.
Responsibilities:
Use machine learning techniques to train, debug, and evaluate models for customer deliveries ranging from quick prototypes to full production-level models.
Perform exploratory data analysis on the large sensory datasets (image, audio, radar) we have gathered, to develop greater understanding of the problem domain.
Define and improve best practices of ML systems development, testing and evaluation.
Work closely with data collection teams, Data and MLOps engineers, and Quality Assurance to improve the quality of our datasets and pipelines.
Work with product managers to help integrate the machine learning solutions and deliver on the desired user experience.
Requirements
3+ years experience and knowledge of applying machine learning or statistical techniques to solve real-world problems, ideally within the computer vision or audio domain (e.g., presence detection, speech processing, etc.).
Bachelor s degree in Computer Science, Data Science, Engineering, or a related field
Strong experience using Python and common data and machine learning-related libraries, such as Keras/TensorFlow or Pytorch, numpy, scipy, Pandas, etc.
Proficient experience with C++ and running optimised machine learning-based applications on edge hardware.
Proficiency using the command line including shell scripting, ideally in a Unix-based environment (e.g. Linux, macOS).
Knowledge of basic signal processing techniques.
Preferred Qualifications:
Experience in deploying models to edge hardware, including experience with PyTorch and ONNX and model compression techniques, e.g. quantisation and pruning.
Reside within the Detroit area or nearby, with the ability to work in a hybrid environment and regularly commute to our Detroit office as needed.
Experience using cloud computing platforms, e.g., AWS or Google Cloud Platform desired.
Benefits
- Comprehensive medical benefits coverage, dental plans and vision coverage.
- Health care and dependent care spending accounts.
- Employee and Family Assistance Program (EAP).
- Employee discount programs.
- Retirement plan with a generous company match.
- Generous Paid Time Off, Sick, and Holidays
- Family Leave (Maternity, Paternity)
- Short- and long-term disability.
- Life insurance and accidental death & dismemberment insurance.