Senior Software Engineering Manager with 20+ years of experience in Backend Architecture, DevOps, and Applied AI — specializing in high-performance systems, local LLM deployment, and intelligent automation.https://pedrammotlagh.com/https://pedrammotlagh.com/blog/ollama-local-ai-advancements/https://pedrammotlagh.com/blog/ollama-local-ai-advancements/Explore the massive architectural upgrades in recent Ollama versions, including native structured JSON outputs, parallel model loading, and full-featured local tool calling.Thu, 28 May 2026 00:00:00 GMThttps://pedrammotlagh.com/blog/gemini-multimodal-video-image-breakthroughs/https://pedrammotlagh.com/blog/gemini-multimodal-video-image-breakthroughs/Google's Gemini models have established a new benchmark for multimodal reasoning. We analyze Gemini 1.5 Pro's 2M token context, native video parsing, and high-fidelity Imagen 3 integration.Mon, 25 May 2026 00:00:00 GMThttps://pedrammotlagh.com/blog/open-source-ai-deepseek-llama-qwen/https://pedrammotlagh.com/blog/open-source-ai-deepseek-llama-qwen/The gap between open-source and proprietary AI has closed. A deep dive into DeepSeek's mixture-of-experts (MoE) reasoning models, Meta's Llama 3.x, and Alibaba's Qwen 2.5 family.Wed, 20 May 2026 00:00:00 GMThttps://pedrammotlagh.com/blog/agentic-workflows-multi-agent-systems/https://pedrammotlagh.com/blog/agentic-workflows-multi-agent-systems/Single prompt chat is obsolete. Discover the architecture of agentic workflows—how multi-agent collaboration, self-reflection loops, and tool orchestration are reshaping software development.Fri, 15 May 2026 00:00:00 GMThttps://pedrammotlagh.com/blog/small-language-models-slm-on-edge/https://pedrammotlagh.com/blog/small-language-models-slm-on-edge/High-performance AI is coming to the edge. Exploring Llama 3.2 1B/3B, Microsoft Phi-3/4, and the magic of WebGPU for zero-server-cost local browser inference.Sun, 10 May 2026 00:00:00 GMT