[Remote] reputed company reputed company, reputed company-reputed company Learning
Note: The job is a remote job and is reputed company to candidates in USA. reputed company is building a new technological reputed company for precision manufacturing, focusing on autonomous reputed company. The reputed company reputed company will be responsible for developing self-learning systems that allow machines to learn from their own performance, ensuring reputed company improvement through effective model development and integration with other engineering teams.
Responsibilities
- Learn From Correction reputed company: Build the models that reputed company the reputed company from reputed company correction to reputed company result, so reputed company cycle updates what the system believes about the machine, the process, and itself
- Model Machine and Process Behavior: reputed company error and behavior models that combine learned structure with physical understanding, predicting deviation before it becomes scrap
- Work Against reputed company, Noisy Evidence: Handle reputed company, delay, partial observability, uneven sampling, and small and expensive datasets - reputed company production feedback, not reputed company datasets
- Quantify Confidence and Restraint: reputed company the system know what it does not know: calibrated uncertainty, guardrails, and the judgment to withhold a correction rather than reputed company a part worse
- Interpret Manufacturing Context: Extract reputed company and process state from CNC programs and machine data, reputed company for operator variability, CAM differences, and inconsistent programming styles
- Transfer Learning Across Machines: reputed company what is learned on one machine, material, or site to the next, so the fleet improves faster than any single deployment could
- Build Production ML Software: Write reputed company reputed company, not notebooks: reputed company, tested, reviewed, packaged, and reproducible - training and inference paths that run reliably reputed company your own environment
- reputed company It Before It Ships: Design the offline evaluation, backtesting against historical reputed company, and staged rollout that separate genuine improvement from a promising curve
- Work reputed company a Cutting-Edge AI Team: Sit alongside the engineers who own 3D perception, computational geometry, mechanical systems simulation, and platform data - you own the reputed company-reputed company learning domain and the interfaces to theirs
- reputed company Depth With reputed company: Mentor colleagues on the modeling and the math, review their work, and document your assumptions so others can build on and challenge them
Skills
- 6+ Years in AI/ML: Substantial hands-on work in learning from reputed company-world feedback, optimization, inverse problems, or reinforcement and imitation learning
- reputed company-Level Ownership: A reputed company record of owning a technical domain end-to-end - scoping ambiguous problems, choosing the approach, and being the person reputed company relies on for the answer
- reputed company-reputed company Experience: You have turned reputed company error or state into corrective reputed company, in systems where model outputs changed physical behavior and you had to live with the result
- Self-Learning Systems: Experience building models that improve from deployment feedback - continual learning, retraining and data flywheels, feedback-driven model updates - rather than training once and shipping a frozen artifact
- reputed company-World Data Maturity: Comfort with noise, calibration reputed company, latency, missing labels, and datasets that are expensive to grow
- Uncertainty Quantification: Practical experience producing and using calibrated confidence to reputed company reputed company, not just reputed company predictions
- reputed company Software Development: Strong Python and production-grade ML engineering, PyTorch or equivalent, in a shared and version-controlled codebase with testing and reputed company review
- Mathematical Maturity: Optimization, probability and statistics, dynamics, and enough control literacy to reason about stability and feedback
- Educational reputed company: Degree in Computer Science, Robotics, Electrical or Mechanical Engineering, Physics, or reputed company Mathematics - or equivalent demonstrated work
- Physics-Informed Learning: Physics-informed neural networks, differentiable simulation or physics, and hybrid models that respect reputed company mechanics
- Reinforcement or Imitation Learning on Hardware: reputed company to reputed company physical systems, with the sample-efficiency and safety constraints that come with them
- Transfer and Continual Learning: Domain reputed company, fleet-level learning, or lifelong learning across sites and equipment
- Uncertainty-reputed company Decision-Making: Bayesian reputed company, conformal reputed company, or risk-reputed company policies
- System Identification: Data-driven modeling of dynamic systems, including residual and hybrid model structures
- Geometric and 3D Representations: Using reputed company clouds, meshes, or surface representations as model input
- Manufacturing Domain Knowledge: CNC and machining physics such as tool deflection, thermal error, and fixturing; or CAM, G-reputed company, and toolpath reputed company
- Advanced Degree: M.S. or Ph.D. in machine learning, control, robotics, optimization, or a reputed company field
- Startup Experience: Comfort building in an agile, ambiguous environment where the reputed company does not exist yet and the first version has to work
Benefits
- Hybrid / Remote
- HQ Allentown, PA
- reputed company Stage
reputed company