Jahanzeb R

AI Engineer

SOC Solutions • Full time

Feb 2025 - Present • 1 yr 3 mos

Currently working on end-to-end AI product development, translating research concepts into scalable, production-ready solutions across diverse real-world use cases. My role spans solution architecture, model experimentation, evaluation, optimization, and deployment. I designed and implemented a voice cloning pipeline using advanced speech synthesis and voice modeling techniques to replicate speaker tone, timbre, and speaking style from raw audio inputs. This involved preprocessing, feature extraction, model fine-tuning, and inference optimization for high-fidelity output. I developed an AI-powered personality development assistant leveraging large language models (LLMs) to generate adaptive content, structured learning paths, and goal-oriented guidance. The system integrates prompt engineering, dynamic context handling, and user-personalized response generation. I implemented an early-stage Retrieval-Augmented Generation (RAG) architecture by combining vector embeddings with LLM-based query answering. This improved factual accuracy, contextual relevance, and reduced hallucinations in domain-specific tasks. Additionally, I contributed to building an AI-driven SEO optimization tool for SaaS platforms, generating high-quality, keyword-optimized content to enhance search visibility, improve ranking potential, and support organic growth strategies. Key Achievements:   • Delivered production-ready AI pipelines from prototype to deployment   • Improved content relevance and user personalization through structured LLM workflows   • Enhanced information retrieval accuracy using embedding-based search   • Supported measurable SEO performance improvements through AI-generated copy