I’m looking for a developer who can create an AI-powered calling agent that I can brand, customise, and resell to multiple clients. The core system must support: • Voice personalisation – synthetic voices that can be swapped or fine-tuned per client. • Native CRM integration – at minimum with common REST / webhook endpoints so I can plug the agent into different CRMs without rewriting code. • Flexible call workflows – drag-and-drop or JSON-based logic so I can change greetings, branching questions, and hand-off rules on a per-client basis. • Training from URLs or custom data sources – I should be able to point the agent to a website, PDF set, or proprietary knowledge base and have that material shape answers in real time. My main goal is not a single use-case like customer support or sales; instead, I need a modular foundation I can tailor for whichever scenario a future client needs. Clean, well-documented code and a straightforward deployment pipeline (Docker/Kubernetes or similar) are essential so I can spin up new white-label instances quickly. Acceptance criteria 1. Demo instance showing voice personalisation, CRM webhook, and at least two switchable call flows. 2. Admin UI or config files that let me add a new website/data source and retrain without touching code. 3. Installation guide plus API docs clear enough for a junior developer to follow. 4. Source code handed over in a private Git repository. If you have previously built Twilio, Dialogflow, or custom LLM telephony solutions, please share brief details when you apply.