Project Description
I want to embed an AI-powered assistant inside my home-interior e-commerce site so customers feel they’re chatting with a knowledgeable designer rather than clicking through catalog pages. The agent must be able to:
• Recommend products that fit a customer’s style, dimensions, and constraints
• Upsell complementary items naturally within the same conversation
• Solve user queries in real time (materials, installation, care, warranty, etc.)
• Add chosen items to the cart on command
• Book service appointments (measurements, installation, designer visits)
• Prepare instant quotations and full BOQ, complete with budget breakdowns
This effectively turns the agent into a virtual interior designer and sales consultant.
Scope of work
– Architect the conversational flow, intent detection, and slot-filling so the agent can juggle recommendations, upsells, and support questions without losing context.
– Connect to our product database, pricing engine, cart API, and service-booking calendar.
– Generate dynamic quotations/BOQs in PDF or email from live data.
– Implement a feedback loop so the agent learns from conversions and abandoned carts.
– Deliver a front-end widget (React preferred) plus well-documented backend code that I can deploy on AWS.
Acceptance criteria
1. A shopper can start with an open-ended style question and end with a populated cart and confirmed service slot in a single session.
2. Upsell rate and response quality are logged for A/B testing.
3. Quotation/BOQ exports show correct SKUs, quantities, prices, and discount logic.
4. All personal data handled in compliance with GDPR/CCPA.
If you’ve built retail or design chatbots using LLMs, RAG, or vector search—and you can demonstrate live integrations—let’s talk timeline and milestones.