I want to deploy a fully-functional AI assistant built on a free Chinese open-source model—Qwen 2.5, Minimax01, Kimi 1.5 or Open Manus are all acceptable. Phase one is crystal-clear: the system must start producing qualified leads for short-term business loans, bridge loans, consolidation loans, equipment financing, commercial refinance and purchase, and even grant-seekers. Accuracy, compliance, and the ability to track every interaction matter more to me than flashy extras. Here’s how I picture the first release: • Model integration and fine-tuning on my loan-specific prompts and historical data • A lead-generation workflow that scrapes or ingests approved data sources, segments prospects, and outputs clean CSV or pushes directly into my CRM • Simple dashboards or logs so I can see volumes, sources, and basic performance stats in real time Once this core is humming, we will expand to automated email sequences, follow-up logic, AI-driven cold-calling that routes live prospects, and any revenue-positive side gigs the model can handle. For now, show me the working pipeline, walk me through the code, and prove the leads are real and contactable. Python, LangChain, vector databases, or your preferred stack are fine as long as they remain license-free and easy to maintain. I’ll supply access credentials, sample data, and compliance guidelines the moment we kick off. Acceptance criteria: the system must deliver a batch of fresh, validated leads and document the exact steps so I can rerun or scale it independently. I will pay 50.00