Zaafir Rizwan

I am a AI Engineer specializing in Voice AI, Large Language Models (LLMs), and Retrieval-Augmented Generation (RAG) with a strong focus on building production-grade AI systems. Over the past few years, I’ve worked in fast-paced startup environments where I’ve designed and deployed scalable AI infrastructure across AWS, GCP, and Azure, delivering real-world AI products with sub-500ms latency and high reliability. zaafir_rizwan

My expertise lies in real-time conversational AI, voice agents, and agentic AI systems. I have built end-to-end voice AI pipelines using technologies like LiveKit, WebRTC, Deepgram, ElevenLabs, and Cartesia, integrating them with telephony platforms and enterprise systems to create intelligent voice assistants. I also design advanced RAG architectures, multimodal AI workflows, and LLM-based applications, leveraging models such as Claude, Gemini, and Llama. zaafir_rizwan

Beyond model development, I focus heavily on MLOps and AI infrastructure, using tools like Docker, Kubernetes, OpenTelemetry, ClickHouse, and MLflow to ensure scalable deployment, monitoring, and continuous improvement of AI systems. My work has helped reduce operational costs by up to 40% and accelerate deployment cycles by 60%, while maintaining production reliability. zaafir_rizwan

I’m particularly interested in voice AI, agentic systems, and real-time multimodal AI, and I enjoy building systems that bridge cutting-edge research with practical applications. I’m always excited to collaborate with teams pushing the boundaries of AI-driven products and intelligent automation.