Focus Areas
Applied research lines where we invest our time and compute.
Medical AI & Validation
Research-grade pipelines for medical imaging and biosignals.
✓ Imaging & signal analysis: chest X-ray, ECG, and multi-modal workflows
✓ External validation: distribution shift, robustness, and interpretability
✓ Reproducible protocols: from training to evaluation
Built to hold up under scrutiny, not just on curated benchmarks.
Computer Vision & Spatial AI
Vision systems for detection, segmentation, tracking, and 3D understanding.
✓ Object detection & segmentation: YOLO, SAM, and custom architectures
✓ 3D reconstruction and geometry-aware pipelines
✓ Multi-modal perception: video, depth, and sensor fusion
From controlled experiments to deployable prototypes.
Voice Systems & Real-Time AI
Low-latency audio pipelines and real-time AI systems.
✓ Streaming ASR/TTS and live translation workflows
✓ Tool-using AI workflows for real-world tasks
✓ Optimized inference for consumer-grade GPUs
Designed for latency, continuity, and practical use.
GPU Optimization & Edge Deployment
Making advanced models run efficiently on limited hardware.
✓ GPU memory optimization and throughput tuning
✓ Quantization & accelerated inference: INT8/INT4, GGUF, TensorRT
✓ Edge deployment: Jetson, embedded devices, and on-device AI
Because research matters more when it actually runs.