TFLite Jobs in Computer Vision
Browse CV roles that require TFLite across all industries and experience levels.
0 open positions
Open Positions
What is TFLite?
TensorFlow Lite, now LiteRT, runs neural networks on mobile phones, embedded Linux boards and microcontrollers with a small binary footprint and support for hardware delegates. It has the broadest device coverage of any on-device inference runtime.
Where TFLite is used
Mobile vision features, battery-powered cameras and sensors, and microcontroller-class vision where the entire model must fit in a few hundred kilobytes.
Roles that ask for TFLite
- Mobile AI Engineer
- Edge AI Engineer
- Embedded Vision Engineer
- Android Developer, ML
- ML Systems Engineer
Related skills & tools
TFLite jobs — common questions
What are delegates and why do they matter?
Delegates route parts of the graph to accelerators — GPU, NNAPI, Hexagon DSP, Edge TPU — rather than the CPU. Getting delegate support right often makes the difference between a feature that ships and one that drains the battery.
How small can models get?
TFLite Micro targets microcontrollers with tens to hundreds of kilobytes of memory, running keyword spotting and simple vision models. Getting there requires aggressive quantisation and architectures designed for the constraint.
Does TFLite only work with TensorFlow models?
Primarily, but PyTorch models can reach it via ONNX conversion, though the path is less reliable. Teams targeting TFLite seriously often train in TensorFlow to avoid conversion friction.