AI Virtual Sales Assistant Platform
Budget: $250 – $750 AUD
I want to build an end-to-end platform that lets any business spin up its own virtual sales assistant, train it on brand-specific data, and deploy it in minutes to a website, a mobile app, or an eCommerce storefront such as Shopify. The assistant must converse naturally, answer customer questions, recommend the right products or services, and qualify leads before handing them off to a human team or a CRM.
The project covers everything from the AI core to the customer-facing widgets. I need a secure multi-tenant backend where companies can upload FAQs, product feeds, and sales scripts, then fine-tune a large-language-model–based agent around that content. A no-code dashboard should expose conversation flows, tone controls, and performance analytics. On the delivery side, I expect lightweight JavaScript and mobile SDKs plus a Shopify app so clients can drop the assistant into their storefront without engineering effort.
Key expectations
• Natural language chat powered by a modern LLM (OpenAI, Anthropic, or an equivalent open-source model)
• Real-time product recommendation logic that can pull data from APIs or CSV feeds
• Lead-qualification prompts with scoring that syncs to HubSpot, Salesforce, or a webhook
• REST / GraphQL APIs for integration, OAuth-based authentication, and role-based access control
• Usage and conversion analytics visible in the dashboard
I will provide sample datasets, brand guidelines, and UI wireframes. Deliver the source code (preferably Python for the AI layer and either Node.js or Go for the platform services), database schema, container-ready deployment scripts, and concise documentation so another engineer can extend the system. A short live demo on a staging domain will serve as final acceptance.
The project covers everything from the AI core to the customer-facing widgets. I need a secure multi-tenant backend where companies can upload FAQs, product feeds, and sales scripts, then fine-tune a large-language-model–based agent around that content. A no-code dashboard should expose conversation flows, tone controls, and performance analytics. On the delivery side, I expect lightweight JavaScript and mobile SDKs plus a Shopify app so clients can drop the assistant into their storefront without engineering effort.
Key expectations
• Natural language chat powered by a modern LLM (OpenAI, Anthropic, or an equivalent open-source model)
• Real-time product recommendation logic that can pull data from APIs or CSV feeds
• Lead-qualification prompts with scoring that syncs to HubSpot, Salesforce, or a webhook
• REST / GraphQL APIs for integration, OAuth-based authentication, and role-based access control
• Usage and conversion analytics visible in the dashboard
I will provide sample datasets, brand guidelines, and UI wireframes. Deliver the source code (preferably Python for the AI layer and either Node.js or Go for the platform services), database schema, container-ready deployment scripts, and concise documentation so another engineer can extend the system. A short live demo on a staging domain will serve as final acceptance.