Lead AI Engineer
Digital Green14 days ago
Bengaluru, IndiaStaff+
Responsibilities
- Build and evaluate multilingual ASR systems for agricultural content and rural speech patterns using systems such as Whisper, Azure Cognitive Services, Speechmatics, and Wav2Vec-based models.
- Develop audio pipelines for transcription, diarization, noise handling, accent robustness, and field-recording validation.
- Develop computer vision models and image/video pipelines for crop disease and pest detection, animal identification, livestock health assessment, and behavior monitoring.
- Create and optimize production inference pipelines for multimodal image and audio models, including low-light, blur, noise, and varying-camera-angle conditions.
- Define evaluation frameworks, benchmarks, monitoring, and continuous-improvement processes for vision and speech models.
- Test models against adversarial inputs, spoofing, data leakage, and other security and safety risks while implementing input validation, content filtering, and anomaly detection.
- Translate product requirements into model tasks and evaluation strategies, mentor junior engineers, and collaborate with backend, DevOps, security, product, and field teams.
Requirements
- Bachelor's, Master's, or Ph.D. degree in Computer Science, Engineering, or a related field focused on Machine Learning, Computer Vision, or Signal Processing.
- At least 6 years of experience as an AI/ML engineer, data scientist, or software engineer, including hands-on machine learning model deployment.
- Deep expertise in computer vision and speech/audio processing pipelines from data preprocessing through production deployment.
- Experience with ASR models, multilingual audio processing, and computer vision classification, detection, and enhancement tasks.
- Proficiency with Python and computer vision/audio frameworks such as PyTorch, OpenCV, torchaudio, and Librosa.
- Strong understanding of evaluation harnesses, benchmarking, and QA tooling for perception and speech models.
- Experience with model quantization, pruning, and distillation for on-device or edge deployment.
- Strong interpersonal and communication skills and the ability to work collaboratively.
- Preferred qualifications include experience building evaluation, testing, or data-pipeline tools; open-source contributions; and model deployment with Triton, TorchServe, TensorRT, or similar technologies.
Benefits
- Competitive and comprehensive compensation and benefits package.
- Opportunity to work on AI systems with direct social impact for smallholder farmers.
- Ownership of end-to-end AI systems from research and prototyping through production deployment at scale.
- Collaboration with a mission-driven, multidisciplinary team and leading philanthropic partners.
Categories