I am a Senior Product Manager & Developer with 18 years of experience building robust, scalable, and high-performance architectures.

Currently, my main focus is on developing DeepTech, Edge AI, and Multimodal AI solutions. I build offline and hybrid pipelines that run local computer vision processing and intelligent voice agents at the edge (Edge Computing), combining dedicated hardware, low latency, and full data privacy.

• 🚀 Passionate about optimizing computer vision models to run efficiently on low-power edge hardware.
• 🎙️ Designing AI voice agents that communicate with near-human latency.
• 🤖 Firm believer in leveraging agentic workflows to automate complex enterprise integrations.

• 🎙️ Voice Agents & Telephony (Conversational AI): Developing ultra-low latency intelligent voice agents using modern frameworks like LiveKit, Pipecat, WebRTC, and native telephony integrations via SIP.
• 👁️ Edge Computer Vision (Edge AI): Implementing and optimizing real-time object detection models (YOLO / YOLOv26) running directly on dedicated hardware silicon (NVIDIA Jetson AGX Orin and reComputer RK). Model conversion and acceleration using ONNX, CUDA, and TensorRT.
• 🧠 Multimodal AI & Local Models (Local LLM / RAG): Configuring local AI pipelines with Ollama (offline LLMs), Whisper for real-time transcription, and advanced Agentic RAG architectures.
• 🏠 0-to-1 SaaS & Real Estate Tech: Building SaaS platforms for the real estate market from scratch, leveraging rapid prototyping and direct user validation to iterate fast, refine the product, and deliver maximum value.

Here are my up-to-date development statistics. (Note: this includes my contributions to private repositories, where most of the heavy edge engineering and commercial development happens).

I'm always open to discussing new architectures, edge engineering, multimodal AI, or complex integration challenges!

_{“The best way to predict the future is to invent it.” — Alan Kay}