Budget: 11995 UAH Deadline: 10 days
Hello! I see the task "Creating an AI assistant for communication with Clients." I have experience with PHP, Laravel, WordPress, OpenCart, HTML/CSS/JavaScript, Node.js, API integrations, and refining online stores. I can quickly understand the current state, carefully implement changes, and deliver a clear result. I am ready to discuss the details and start working.
Budget: 6000 UAH Deadline: 6 days
Good day, Yulia!
We create turnkey solutions — a chat assistant for the website that responds based on your products and settings and correctly forwards complex queries to the manager.
Here’s how I see the implementation:
• we embed the chat widget into your website under the existing code, without changing the platform;
• a knowledge base from your product descriptions, FAQs, and instructions — the bot responds only from it, without fabrications;
• if a question is outside the base or requires a manager, the bot formulates a request and forwards it to your channel (email, Telegram, or CRM — let me know what’s convenient).
Regarding the platform: with 100–300 queries per month, the cost of the model is negligible on any, so you should choose based on the quality of Ukrainian-language responses and working with the base. I would recommend GPT-4-class or Claude for dialogue accuracy, Gemini Flash as a budget option. Before starting, I will provide a brief comparison specifically for your case.
I am taking it at the starting price — I am gathering initial reviews here, so I am making it advantageous. I am ready to show a live demo so you can test the flow to the solution.
One question: where should the escalation to the manager go — email, Telegram, or CRM?
Budget: 5000 UAH Deadline: 4 days
Hello! I have created similar AI assistants (widget-helper in the corner of the screen). I can implement it turnkey for your custom website.
I will connect the chat so that it responds exclusively about your products, capabilities, and settings. For this, I need your materials: website, FAQ, price lists, or just a file with typical responses from managers. I will set up strict limitations (RAG technology) so that the bot does not make up answers — if there is no information or the request is non-standard, it will correctly pass the dialogue to a human.
Regarding your question about models: with a volume of 100-300 requests per month, the cost of tokens in Gemini, GPT, and Claude will be minimal (within $2-5 per month). Therefore, the choice should be purely based on quality. I would recommend GPT-4o-mini or Claude 3.5 Sonnet — they currently work best with Ukrainian and maintain a strong knowledge base. Gemini sometimes struggles to adhere to strict instructions in such tasks.
Since your website is on custom code, I will prepare a lightweight script for the widget and connect everything via API, so as not to disrupt the current operation of the site.
I will deliver a working version in 4 days, at a cost of 5,000 UAH.
Please let me know where it would be most convenient for your managers to receive escalated dialogues from the bot: in Telegram, by email, or directly in CRM?
Budget: 15000 UAH Deadline: 7 days
Hello! I specialize in developing commercial AI agents and backend integrations both through n8n and scripts, so I can implement this assistant "turnkey" for your website. I have extensive experience in developing commercial AI agents (I created an agent that completely replaced the first line of managers and recorded data in CRM systems for a waterproofing and insulation company - examples are available). I am ready to discuss your knowledge base and functionality in more detail during a short call or in text format.
Budget: 5000 UAH Deadline: 5 days
Hello, Yulia!
Regarding the choice of platform — I will provide a specific expert assessment as requested.
For 100-300 queries per month, the cost of tokens for all three models is negligible (≈$2-8/month), so the choice should be based on the quality of responses rather than price.
My conclusion after daily work with the Claude API in production projects:
→ Claude Sonnet — the best choice for customer support. It handles instructions better than others, does not "fantasize" outside the knowledge base, and correctly escalates to a manager.
→ GPT-4o — good, but often "fills in" answers when there is no information in the base.
→ Gemini — weaker in working with custom instructions.
How I will implement it:
1. Chat widget
I will integrate it into your existing code without rewriting the frontend — a small JS script + REST API backend on FastAPI.
2. Knowledge base (RAG)
You upload documents/FAQs/product descriptions → automatically split into chunks → vector storage (ChromaDB) → Claude responds strictly based on your data.
3. Escalation to manager
If confidence is low or the client requests a manager → notification in Telegram or via email with the context of the dialogue.
4. Logging dialogues
All conversations are saved in the database — you can see frequent questions and improve the knowledge base.
Stack: FastAPI + Claude API + ChromaDB + PostgreSQL
Question: in what format is the product information currently stored (website, PDF, Google Docs, spreadsheet)? And where is it more convenient to receive escalations — Telegram or email?
Budget: 24000 UAH Deadline: 14 days
We have a practically ready solution for such an AI assistant, which can be quickly adapted to your website, products, and request transfer scenarios to the manager. ))
Estimated timeline - 10-14 days for the first working version.
Regarding the budget - starting from 24,000 UAH for launching a basic assistant with a knowledge base, integration into the website, and transferring complex requests to the manager.
I would advise not to focus solely on Gemini, GPT, or Claude by name. It’s better to compare 2 models on your real questions and choose the one with better answer quality and price for 100-300 inquiries per month. For such a volume of requests, the costs for the model will likely be small; the main work involves structuring knowledge, scenarios, protection against fabricated answers, and transferring to the manager.
In implementation, I see it as follows:
- Collection and preparation of materials on products
- Chat widget for your website on custom code
- Searching for answers in the knowledge base
- If data is unavailable - careful transfer of the question to the manager
- Dialogue log to see frequent questions and improve the assistant
Two points need clarification:
- Where is the product information currently stored - website, files, spreadsheets, CRM, instructions?
- Where exactly should the request be sent to the manager - admin panel, email, Telegram, CRM, or another channel? We will agree on this technically within the marketplace.
Examples of similar works:
- https://business.ingello.com/fractal - process automation with agency logic
- https://business.ingello.com/vorfahr - AI and automation for business processes
- https://business.ingello.com/tts - working with AI communication and voice scenarios
Main page for our system solutions for freelancehunt:
https://systems-fl.ingello.com/ua
I am ready to discuss the details here on the marketplace; I am available (:.
Budget: 4000 UAH Deadline: 14 days
Good day!
I am interested in your project — here is my profile. I have already developed and launched a chatbot with LLM integration (Python + Gemini API, works 24/7), so the task is clear to me.
What I propose:
To implement an assistant based on Google's Gemini — for your volume (100–300 requests/month) this is the optimal option: Gemini has a free/low-cost plan that comfortably covers this volume, so your operating costs will be minimal. The bot will answer questions about the product/settings based on your information, and when a request is outside its competence or requires a manager, it will pass it on.
Budget: 10000 UAH Deadline: 5 days
I read the task. An AI assistant is needed in the chat window on your website: it answers questions about products, settings, and capabilities, and when questions are outside the knowledge base or require a manager, it forwards the request to the manager.
I will provide a specific recommendation for the platform. For 100-300 queries per month, it doesn't make sense to maintain a separate heavy infrastructure: we will take one of the models (Claude, GPT, or Gemini), on top of which we will implement RAG based on your knowledge base, so the assistant answers strictly from your materials and does not make things up. I will compare the models based on quality in Ukrainian, price for such volume, and latency, and provide a brief justification for the choice. Escalation to the manager will be triggered: if confidence is low or the question is outside the base, the assistant will not speculate but will pass the dialogue.
I will integrate with your own website code through a widget plus a backend endpoint, without rewriting your frontend.
I have created a voice AI assistant with RAG search based on a knowledge base for a real business, so the logic of passing to the manager and responses strictly from the base is familiar to me. Please clarify in what format your knowledge base is currently (website, documents, spreadsheet), so I can estimate the volume of data preparation.
Budget: 3000 UAH Deadline: 5 days
Hello, Yulia! I specialize in AI assistants for customer communication — I have built my own AI-CRM (Mercon), where the agent responds to customers about products, availability, and conditions, while complex inquiries are forwarded to a human.
I will create for your website:
— a chat widget with an AI assistant (GPT / Gemini / Claude — we will choose the optimal one);
— responses about products, settings, and capabilities — based on your knowledge base, without fabrications;
— when information is lacking or the inquiry is complex — a proper transfer to an operator, not a "hallucination";
— tone and style tailored to your brand.
I work with prompt engineering daily, so the assistant will respond consistently and accurately.
I am ready for a short call to clarify the sources of information and the scenarios for transferring to a human. Approximately 5 days.
Matviy
Budget: 1000 UAH Deadline: 1 day
Hello, I have commercial experience in developing agents. If needed, I can provide examples. I'm interested in your offer; let's discuss the details in private.
Budget: 10000 UAH Deadline: 3 days
Hello! I would like to propose creating an AI assistant based on GPT or Claude integrated into your website, which will be able to answer questions about products, settings, and service capabilities, and in case of insufficient information, automatically transfer the dialogue to a manager. In similar projects, I have implemented AI consultants for client qualification through Chatfuel and Voiceflow, as well as built automated scenarios for handling inquiries through Make.com. This allowed reducing the load on support and providing quick responses 24/7. For your volume (100–300 inquiries per month), I would recommend GPT or Claude due to better response quality and working with a knowledge base. Gemini can also be considered, but the choice depends on the structure of the content that the assistant will use. For an accurate estimate, I would like to clarify: do you already have a prepared knowledge base (FAQ, documentation, instructions) on which the assistant should be trained?
Budget: 25000 UAH Deadline: 1 day
In general, there are already ready-made solutions that can be trained on your data, but you can also create your own backend with a widget. What do you think would be better? I can respond in private messages.
Budget: 26985 UAH Deadline: 30 days
Good day, we have been specializing in website development and promotion since 2018. We have extensive experience working with various websites and online stores. I suggest filling out the brief to create a commercial proposal for you. Brief for website development - https://lionex.com.ua/s/brief-na-rozrobku-saitu Online store - https://lionex.com.ua/s/brief-na-internet-magazyn Landing page - https://lionex.com.ua/s/brief-na-lending SEO - https://lionex.com.ua/s/brief-na-seo-audyt We also offer a cool platform for quick text rewriting - https://textory.com.ua/go/kFoMHu3N You can also check out our work on the website.) If you have any questions, feel free to reach out, and we will gladly provide consultation.
Budget: 10000 UAH Deadline: 3 days
Good day, Yulia. I have experience integrating the platform. The client actively uses it and is very satisfied. Please write to me personally, and I will tell you what the platform is, how much time is needed for integration, pricing, and so on.
Budget: 2000 UAH Deadline: 2 days
Good day!
AI assistants for customer support are my main topic, so the task is clear. I will describe how I see it and immediately provide an expert assessment of the platform you requested.
I will create a chat widget for your website (it’s on custom code — I will integrate it without problems), and under the hood — an assistant based on RAG: it answers questions about products, settings, and capabilities strictly based on your knowledge base, and if a question is outside the base or "for the manager only" — it honestly directs the request to the manager instead of making up an answer. Grounding is key here: it’s bad for customer support when a bot fantasizes, so I will set it up to rely only on real information.
Regarding the platform (expertly): with 100-300 requests/month, the cost in Gemini, GPT, and Claude is minimal — so you should choose based on the quality of answers, not the price. I would recommend Claude (Sonnet) or GPT — they best adhere to instructions and respond strictly based on the base, without "fantasies," which is critical for support. Gemini Flash is the cheapest, but at such a volume, the savings are insignificant. We will finalize the selection during testing — we will run your real questions and compare which one responds more accurately.
The most important factor for quality is the knowledge base (product descriptions, settings, FAQs, typical responses). Please let me know: do you already have it in some form (documents, website pages, tables) or does it need to be gathered and structured? This will affect both the quality of the bot and the scope of work — so we will start from this.
Budget: 27000 UAH Deadline: 30 days
Hello, Yulia!
I have developed many agents with different working mechanics.
The estimated cost of the written code is between $1200 - $1500.
The price mainly depends on the complexity of the knowledge base and related adjustments (product feed for the site and so on).
Please write to me, I will consult, explain in detail how it works, and provide a solution.
If such a budget is not available, we can discuss alternative solutions.
From experience, this will take about a month of work.
In the end, I will deliver a fully functional system without bugs, along with instructions for setup (and future adjustments. I will configure everything for you during the process).
Further support is free forever.
Budget: 10000 UAH Deadline: 10 days
Good day.
Our team has many years of experience in developing ERP, CRM, CMS, and specialized software for businesses. We create effective digital solutions that help automate processes, increase productivity, and scale companies.
We work with modern technologies — from bots and scripts to AI agents and analytical systems. We develop websites of varying complexity. Our portfolio includes implemented ERP solutions for the hotel business, as well as for companies engaged in the import and sale of goods, and our own product XFitness — an ERP system created specifically for fitness clubs.
We are ready to implement your project and offer the best solution tailored to your needs.
Our portfolio: 
Freelancehunt
We specialize in the following areas:
- Development of ERP Systems
- Development of CRM Systems
- Development of Websites of any complexity
- Development of CMS Systems
- Support for Websites
- Development of OpenCart
- Support for OpenCart
- Modification of OpenCart
- Enhancement of OpenCart
- Development of WordPress
- Support for WordPress
- Modification of WordPress
- Enhancement of WordPress
- Development of ECommerce
- Support for ECommerce
- Modification of ECommerce
- Enhancement of ECommerce
- Development of Web Applications
- Support for 1C Servers
- Support for Web Servers
- Development of mobile applications
- Data parsing
- Development of bots
- Development of AI agents
and on the following technologies:
- Python
- PHP
- Laravel
- Symfony
- Yii2
- JS
- NodeJS
- jQuery
- TypeScript
- MySQL
- HTML
- CSS
- Vue
- Nuxt.js
- React
- React Native
- C++
Budget: 18000 UAH Deadline: 12 days
Hello!
I can help with creating an AI assistant for communication with clients on your website.
For the task, I see the optimal MVP format: a chat widget on the site, a backend for working with the AI model, a knowledge base on products/settings/capabilities, as well as the logic for transferring requests to a manager in cases where the bot does not have enough information or the question requires a manual response.
We work with AI tools, chatbots, integrations, and internal business automations. It is important to create not just a "chat with AI," but a controlled assistant that responds only based on the provided information, does not invent answers, and correctly forwards complex inquiries to the manager.
What I can implement:
* chat widget for the site;
* connection of the AI model;
* knowledge base on products;
* response scenarios;
* fallback to the manager;
* logging of dialogues/requests;
* basic user instructions;
* recommendation on the model: GPT, Gemini, or Claude after assessing your data and examples of questions.
Approximately for the MVP: 18,000 UAH, timeframe 12 days.
After reviewing the website, product structure, and examples of typical questions, I will be able to confirm the scope more accurately and suggest the best model for your budget and quality of responses.
Our website: https://sortitout.work/
Budget: 15700 UAH Deadline: 8 days
Your website needs an assistant that accurately answers product questions and smoothly escalates complex inquiries to a manager—without gaps in responses and without overloading the team.
I have implemented similar chatbot integrations for e-commerce and B2B websites: setting up a knowledge base, escalation logic to a human, and web integrations with custom code. With a volume of 100–300 inquiries per month, the optimal choice is Claude API (Anthropic) or GPT-4o Mini—both provide quality responses at minimal costs (approximately $2–8/month for tokens); Gemini should be considered if you already have a Google Cloud infrastructure.
Implementation includes: preparing a knowledge base for your products in a structured format, setting up a system prompt with clear escalation rules, integrating a chat widget into your custom site via REST API, and a mechanism for passing inquiries to the manager (email / webhook / CRM—depending on your stack). To better assess the scope: in what format is the product information currently stored (documents, website pages, spreadsheets) and which channel is prioritized for receiving inquiries from the manager?
Estimated timeline—5–8 days, cost starting from $350 depending on the volume of the knowledge base and the complexity of the integration.
Budget: 4999 UAH Deadline: 5 days
Good day, Yulia!
Such solutions are my specialty.
Regarding the platform: for 100-300 requests per month, I recommend the Claude API (Anthropic) — the best quality of Ukrainian language responses, accuracy regarding products, and correct transfer to the manager. The cost will be minimal at this volume.
What I will do:
- Chat widget embedded in your website
- Knowledge base with your products and FAQ
- Bot responds only from the database — no fabrications
- Unknown request → transfer to the manager (email or Telegram)
- Custom code, without third-party platforms
Timeline: 3-5 days
Cost: 5000 UAH
Where should the escalation to the manager go — email or Telegram?
Budget: 1000 UAH Deadline: 1 day
Good day, Yulia.
Your AI assistant will accept customer inquiries about products and settings through a widget on the website, and complex or unknown inquiries will automatically escalate to a manager — transferring the context of the conversation so the customer does not have to repeat everything.
For 100–300 inquiries per month, the optimal choice is GPT-4o-mini as the main model (low token cost, response speed) with escalation to GPT-4o or Claude 4.6 Sonnet for complex technical inquiries. This will keep monthly expenses within $5–15 instead of $50+ if only top models are used. Gemini should only be considered if multilingual support with Ukrainian is critically important to you — but it falls short in the quality of instructions.
The reliability of responses is ensured through Self-RAG with factual verification: a reviewing agent checks the generated response against your product knowledge base and, in case of discrepancies, returns a token for regeneration. The knowledge base in `pgvector` is protected from Prompt Injection — input data passes through an LLM-validator with a strict JSON-schema output. Instead of passing raw dialogue, each step is governed by a State Machine with a structured JSON state (`{current_topic, product_context, escalation_needed}`), which prevents Subject Drift and blurring of the model's focus.
Orchestration in Go with `goroutines` allows for parallel processing of inquiries without delays, and a hybrid model scheme (cheap for intent classification, expensive only for the final response) minimizes token costs.
The result is faster 24/7 support and significant relief for your team from routine questions.
Stages:
- Requirement analysis and AI platform selection
- Development of the knowledge base and RAG setup
- Integration of the chat widget on the website and API configuration
- Setting up the escalation system to the manager
- Testing, calibrating responses, and optimization
- Launch and documentation
Budget: 16000 UAH Deadline: 5 days
Good day!
I am interested in your project. For your scenario (100–300 inquiries per month, own website, product consultations, and transferring complex requests to a manager), it is quite feasible to create an AI assistant without unnecessary costs and complicated infrastructure.
I can help not only with implementation but also with choosing the optimal platform. At the moment, I would not recommend selecting a model solely based on popularity — it is essential to look at the quality of responses, cost of inquiries, integration, and the ability to work with your knowledge base.
Preliminarily, I see the solution as follows:
• chat widget on the website;
• knowledge base based on your documentation, FAQs, product descriptions, and instructions;
• AI responds only with verified information;
• if the confidence of the response is low or the question goes beyond the knowledge base — automatic transfer of the inquiry to the manager;
• logging of dialogues and the ability to analyze customer inquiries.
Regarding models:
• GPT — currently the best balance of response quality and integrations;
• Gemini — a good option for deep integration with Google services;
• Claude — very strong in handling large volumes of text and documentation.
After familiarizing myself with the product and website, I will be able to provide a reasoned recommendation specifically for your case.
To assess the cost and timelines, I would like to clarify:
— what language should the assistant operate in (UA/EN/other)?
— is there an existing knowledge base (PDFs, instructions, articles)?
— how should the manager receive complex requests (email, Telegram, CRM)?
— is a history of dialogues needed in the administrator's cabinet?
I am ready to propose the architecture of the solution and implement the full cycle — from model selection to launch on the website.
Budget: 20000 UAH Deadline: 7 days
Hello, I have extensive experience with similar AI bots, examples are https://neurofoto.de/ and https://orban-containerdienst.de/. I would be happy to do everything for you in the best way possible.
Budget: 25000 UAH Deadline: 14 days
We have experience in implementing LLM solutions for support automation. We achieve this through integration with the OpenAI API and configuring a knowledge base for your product. The system will automatically route complex inquiries to managers. This will ensure quick responses to customers and reduce the workload on the team.
Budget: 12345 UAH Deadline: 3 days
Hello. Are you planning to integrate the assistant with your product database, or should it respond based on general information?
Details regarding timelines and budget will be clarified in personal correspondence.
Here’s how I propose to execute this project:
1. I will conduct an expert evaluation of Gemini, GPT, and Claude to choose the optimal platform for your 100-300 queries per month.
2. I will set up the AI assistant with clear instructions for answering product-related questions and redirecting complex queries to a manager.
3. I will integrate the chat window into your own website code and test the bot's functionality.
Thank you for considering my proposal. I am excited about the possibility of working with you!
Budget: 1000 UAH Deadline: 1 day
Good day, I am ready to complete your task quickly and efficiently. I have extensive experience in creating various bots. Please write to me in private messages to discuss the details. I would be happy to help :)
