• Projects 31
  • Rating -
  • Rating 655

Budget: 9500 UAH Deadline: 6 days

Good day, Irina!

I have experience in integrating AI models. I am interested in your project and would be happy to help in the development of this widget.

You can familiarize yourself with examples of my work in my profile, and we can discuss the details in private messages.

I look forward to collaborating!

  • Projects -
  • Rating -
  • Rating 435

Budget: 8500 UAH Deadline: 10 days

Good day, Irina!

I can create and set up an AI consultant for a product selection website, implementing the logic in n8n with integration of Google Sheets, chat widget, and OpenAI.

What I can do:

1. Build a chain of clarifying questions for accurate selection among ~10,000 products;
2. Set up search in Google Sheets considering attributes;
3. Generate a response with 5 relevant products, an explanation of the choice, and links to "add to cart";
4. Provide flexible management of logic and scalability;

  • Projects -
  • Rating -
  • Rating 302

Budget: 8000 UAH Deadline: 5 days

Good day! I can create an AI chat widget consultant for the website through n8n. I have already worked with similar logic: Webhook → OpenAI (request determination) → search in Google Sheets → output of 5 products + “add to cart” links.

What I will do:
- I will set up clarifying questions so that the client can accurately find the desired product
- I will optimize the search in a database of ~10,000 products
- The responses will be: 5 products + why exactly these + buttons/links
- Everything can be easily expanded

I would like to clarify a couple of points:
1) On which website will the widget be (Crisp / Tawk / Chatwoot)?

  • Projects -
  • Rating -
  • Rating 171

Budget: 15000 UAH Deadline: 10 days

Hello! Interesting task. I will say right away: direct search in Google Sheets for 10,000 products will work too slowly for chat.

I propose a more modern solution in Python (FastAPI + LangChain):

Search: I will use a lightweight vector database (FAISS or Pinecone). This will allow AI to find products by content (for example, "comfortable sneakers" will find exactly those), rather than just by keywords. And it will be instant.

Logic: Python script instead of complex schemes in n8n. This is more stable and cheaper to maintain.

Integration: I will connect to Crisp or Tawk.to, as you planned.

  • Projects 13
  • Rating 4.9
  • Rating 6 949

Budget: 12000 UAH Deadline: 5 days

Hello, I have experience in implementing similar systems.

In order for the consultant to work instantly and "intelligently," I suggest using OpenSearch (or Elasticsearch) as a vector database.

How I see the implementation:
OpenSearch for speed: Instead of forcing AI to browse 10,000 rows in a table, we will load products into OpenSearch. This will allow for semantic search — the bot will understand the context (for example, that "warm shoes" means boots), even if the word "warm" is not in the title.

Hybrid logic in n8n:

Step 1: n8n receives a request via Webhook from the chat widget.

  • Projects 19
  • Rating 5.0
  • Rating 19 982

Budget: 20000 UAH Deadline: 5 days

Have you ever thought about how to help clients choose the perfect product from 10,000 faster and more accurately? I specialize in creating intelligent chat widgets, and my solutions are always results-oriented. I will propose a logic of clarifying questions that will effectively narrow down the selection. With experience in integrating OpenAI and optimizing processes, I know exactly how to achieve a smooth and clear interaction with the client. By preparing real examples from other projects, I am ready to dive into your needs and together create an AI consultant that will significantly improve customer service. Shall we discuss the details?

Android agent

Proposals concealed

The list does not show proposals concealed by the client or freelancer with a Plus profile, as well as proposals violating rules