AI Engineer (Audio & Edge AI – Savearth)
Location: Hybrid
Employment Type: Full-time
About our Project
Savearth is a technology-driven sustainability solution focused on reducing water consumption in hotels through intelligent devices powered by sound analysis and AI.
Our system uses audio signals to estimate water flow and user behavior in real time, combining embedded systems with machine learning models deployed on edge devices.
Responsibilities
- Design, build, and maintain machine learning models for audio analysis
- Develop and train CNN models for sound classification and regression
- Build and manage audio datasets, including data collection, labeling, and
validation
- Process audio signals into meaningful features such as spectrograms,
mel-spectrograms, and MFCCs
- Optimize models for edge deployment (TFLite / int8 quantization)
- Evaluate model performance and continuously improve accuracy and
robustness
- Work closely with firmware and embedded teams to integrate models into
devices
- Develop data pipelines for training, validation, and inference testing
- Analyze real-world data collected from deployed devices and iterate on
models
Requirements
- Strong experience in Machine Learning/Deep Learning
- Proven experience with CNNs for audio or signal processing tasks
Experience with audio processing techniques (spectrograms, MFCCs, FFT,
filtering)
- Experience with Python and ML frameworks (TensorFlow/PyTorch)
- Experience with data preprocessing and dataset construction
- Understanding of regression models and time-series data
- Experience evaluating model performance (metrics, validation strategies)
- Ability to work with real-world noisy data and improve model robustness
- Availability to travel
Nice to Have
- Experience with Edge AI/TinyML
- Experience with TensorFlow Lite (TFLite) and model quantization (int8)
- Experience working with embedded systems constraints (memory, latency,
CPU)
- Familiarity with ESP32 or similar microcontrollers
- Familiarity with C and C++
- Experience with audio-based IoT applications
- Knowledge of signal processing theory
- Experience with cloud pipelines (AWS, data ingestion, analytics)
What We Offer
- Opportunity to work on a real-world AI product with environmental impact
- Unique technical challenge combining audio, AI, and IoT edge computing
- Collaboration with hardware, firmware, and product teams
How to Apply
* Please send your CV and any relevant portfolio or GitHub links to:rh@savearth.io