
2
2
Jordan
2+ years of experience
I have extensive experience in business and commerce, with nearly nine years in sales, sales management, and strategic advisory roles. My most recent position was as an electric-vehicle consultant, where I developed a global benchmarking standard for evaluating EV performance and adoption. Despite being only 21, I have accumulated deep, practical expertise shaped by real market dynamics, and I’ve read extensively across business, startup strategy, and innovation literature. Alongside my business background, I have strong technical capabilities. I am a full-stack engineer specializing in Generative AI, RAG systems, AI automation, and end-to-end web and mobile development. My skill set includes building scalable backend systems, intelligent AI pipelines, production-ready mobile apps, and deploying full applications on cloud platforms like AWS (EC2, ECR, and containerized services). This combination of business insight and technical execution enables me to build products that are not only functional, but strategically aligned with real market needs—bridging the gap between technology, entrepreneurship, and practical decision-making.

is a domain-specific large language model designed to act as an expert consultant for startups . Developed for the AMD Developer Hackathon, the project solves the problem of generic, "wiki-style" advice generated by standard AI models by delivering immediate, high-impact guidance . The model was built by distilling core principles from renowned Silicon Valley business literature (such as The Lean Startup, Zero to One, and Blitzscaling) into a meticulously engineered dataset of 1,564 ChatML instruction-following pairs . To ensure tactical execution, the model adheres to a strict 4-pillar response architecture: Direct Actionable Advice, Why This Matters, Real-World Examples (grounded in 2025 context), and 'Avoid' variables . Built on the Qwen2.5-7B-Instruct base model, the fine-tuning process was hyper-efficient. By leveraging Unsloth optimization and QLoRA 4-bit quantization, the 7-billion parameter model was trained in under 30 minutes using only 5.5 GB of VRAM . Crucially, the model architecture and LoRA adapters are 100% compatible with AMD's ROCm ecosystem (including AMD Instinct™ GPUs and Ryzen™ AI-powered PCs) for high-performance production inference . The project is accessible via a premium, fully interactive Gradio chat interface deployed on Hugging Face Spaces
10 May 2026

Mr WorkFlow AI is a multi-modal, stage-aware AI mentor designed to support founders and startup teams throughout the entire innovation lifecycle. The platform seamlessly integrates Retrieval-Augmented Generation (RAG), Gemini Vision, and advanced document intelligence to deliver deeply contextual, reliable, and actionable insights. Users can ask open-ended questions, upload pitch deck slides for automated critique, or submit full PDF documents such as business plans, market research, or investor reports for in-depth analysis and structured guidance. What makes the system unique is its ability to automatically classify each query into one or more of the eight core startup stages, ranging from early ideation and problem validation to scaling, operations, and fundraising. Mr WorkFlow AI retrieves targeted recommendations from a finely tuned knowledge base built on thousands of high-quality, structured startup principles and advice vectors inspired by real-world founders, modern AI/SaaS companies, and widely respected entrepreneurship frameworks. The result is a fast, highly accurate, expert-level assistant that serves early founders, experienced entrepreneurs, students, incubators, accelerators, and hackathon teams who need on-demand, dependable strategic guidance. The platform dramatically reduces decision-making uncertainty, accelerates learning, and empowers teams to move from idea to execution with clarity and confidence.
19 Nov 2025