I build intelligent systems that understand the world through audio, images, and data. With expertise across machine learning, deep learning, and real-time AI, I help companies turn raw inputs—speech, video, documents, sensor streams—into powerful, production-ready applications. What I Do Voice AI Development: ASR, TTS, diarization, voice activity detection, emotion classification, realtime WebSocket audio pipelines, and LLM-based conversational AI systems. Computer Vision: Object detection (YOLOv8/RTDETR), image classification, OCR, document intelligence, multi-object tracking, pose estimation, and custom model training with PyTorch/TensorFlow. Data Science & Machine Learning: Predictive modeling, time-series forecasting, embeddings, RAG, NLP, evaluation pipelines, and end-to-end data workflows. Tech Stack ML/DL: Python, PyTorch, TensorFlow, OpenCV, HuggingFace, Transformers Voice AI: Whisper, NeMo, Deepgram, Silero, WebRTC, VAD, TTS models CV: YOLO, Detectron2, OpenCV, LayoutLM, PaddleOCR Pipelines: FastAPI, Docker, Airflow, MLflow, Redis, Kafka Cloud: AWS, GCP, Azure Tools: Jupyter, CVAT, Label Studio, DVC, W&B What You Get Working With Me Clean, production-grade code Clear communication and frequent updates Models optimized for accuracy, latency, and cost Fast turnaround and reliable delivery A partner who thinks about product and engineering Example Projects Built real-time voice assistant using Whisper + LLM with <200ms latency Developed end-to-end document OCR + structured extraction pipeline Trained custom YOLO models for object detection, tracking, and quality inspection Built multimodal AI for audio–video emotion recognition Automated data pipelines for millions of records with MLflow + DVC If you need a serious engineer who understands the full lifecycle—from data → model → deployment—let’s build something great together. Send me a message anytime. I respond quickly.