Bitrix 24 + Bitrix GPT

Budget: 600 USD Deadline: 7 days

Hello. I will create a complete RAG service: automatic import and re-indexing of the Bitrix24 Knowledge Base, chunking, embeddings, Qdrant/pgvector, responses only based on the uploaded context with links, internal bot through imbot.register and POST /api/v1/ask with Swagger/Postman. I have worked with API integrations, webhooks, and RAG pipelines. I can start immediately. On the first day, the import of one article and search with a returned link will be operational. The cost is 600 USD, and the timeframe is 7 days. Is Bitrix24 cloud-based or on-premises?

Budget: 1800 USD Deadline: 18 days

Look, this task is not about a simple bot, but about a separate RAG service between Bitrix24, GPT, a vector database, and external integrations. The budget of 600 USD for the full scope is likely too small - we can simplify and start with a working MVP for 1800 USD and 18 days, and then expand roles, analytics, and search quality.

For implementation, I would proceed as follows - import the Bitrix24 knowledge base via REST API or webhooks, proper chunking, embeddings, vector storage, a GPT response layer only based on the found context, links to source articles, an internal chatbot via imbot.register, external POST /api/v1/ask, and documentation for the Telegram bot developer.

I will clarify 2 points to avoid guessing:
> Is the knowledge base a single common one, or do we need to consider access rights for employees and departments in Bitrix24?
> Is GPT needed through the OpenAI API, Bitrix GPT, or can we choose a model based on price and quality?

Similar projects by Ingello:
> https://business.ingello.com/fractal - AI automation and agent architecture for complex processes
> https://business.ingello.com/vorfahr - AI/SaaS approach, integrations, and automation around data
> https://business.ingello.com/forma-crm - corporate logic, CRM processes, and sustainable internal systems

The main profile for such tasks on the marketplace:
> https://systems-fl.ingello.com Expand

Budget: 600 USD Deadline: 7 days

Rustam, there is an important nuance regarding the chatbot: imbot.register, which others refer to in bets, has been officially marked by Bitrix24 as discontinued in development; the current method is imbot.v2.Bot.register. I will use it right away so that the bot does not remain unsupported.

For 600 USD, the entire scope from the technical specification, the bot in the portal, external API, documentation, and regular re-indexing cannot be gathered; I agree with those who honestly wrote about this in the bets. I propose to close the foundation in the first stage: I will retrieve articles from the Knowledge Base via REST API, break them into chunks, create embeddings, and vector search, while GPT will respond strictly based on the found context with a reference to the source article. The bot will be inside the portal, and the external API for Telegram and documentation will be done as a separate stage after the search in your actual database proves to be functional. The timeframe is 7 days, the cost is 600 USD, and I will confirm the exact estimate when I see the volume of the database and whether multilingual support is needed.

Budget: 1500 USD Deadline: 10 days

Hello! I will complete your task quickly and efficiently.

My recent works
https://indexfast.pro - fast website indexing
https://mono-bank.pp.ua - everything about Monobank
https://mamamia.pp.ua - online store
https://programist.pp.ua/ua/portfolio/ - portfolio of works
https://monitortest.pp.ua - monitor testing
https://keytest.pp.ua - keyboard testing
https://pctest.pp.ua - computer testing
https://devroman.dev/portfolio/ - portfolio
https://vseetut.matviy.pp.ua - layout
https://indexfast.pp.ua - fast website indexing

My portfolio: https://freelancehunt.com/ua/freelancer/romas6ka.html#portfolio
Message me, I will start working today. I will be glad to collaborate with you! Expand

Budget: 600 USD Deadline: 6 days

Good day, Rustem!

The technical task is clear: a RAG service for the Bitrix24 Knowledge Base is needed — with article indexing, vector search, generating answers through GPT without hallucinations (only based on the uploaded context, with a reference to the source), plus a chat-bot inside Bitrix24 and an external API for integration with the existing Telegram bot.

The stack I will use:

Python + FastAPI — service layer, external REST API (POST /api/v1/ask)
Bitrix24 REST API/Webhooks — import and regular re-indexing of articles from the Knowledge Base
Vector DB (Qdrant or similar) — storing embeddings, chunking text
GPT API — generating answers based solely on the uploaded context, mandatory reference to the source article
imbot.register (Bitrix24) — internal chat-bot, processing personal messages and mentions in portal chats
Swagger/Postman — API documentation for the Telegram bot developer

I would like to clarify two points for a more accurate estimate:

Estimated volume of the Knowledge Base (how many articles) — affects the time for initial indexing and the choice of vector DB.
Which LLM is planned — own OpenAI key/other provider, or do we need to find an option within the budget.

Estimated breakdown by stages (rate $15/hour):

Setting up access to the Bitrix24 Knowledge Base, chunking, and initial indexing — 1 day, $75
Vector DB + automatic re-indexing pipeline — 1 day, $75
Integration of GPT: generating answers with a reference to the source, without hallucinations — 1 day, $75
Chat-bot inside Bitrix24 (imbot.register, personal messages, mentions in chats) — 1 day, $75
External REST API endpoint + Swagger/Postman documentation — 0.75 days, $60
Testing and deployment — 0.5 days, $40

Total: 5.25 days of work + 1 day in reserve, $400

Ready to start with clarifying the volume of the Knowledge Base and choosing the LLM provider. Expand

Budget: 600 USD Deadline: 7 days

Good day, Rustam!

My name is Vladyslav Kushnarov. I specialize in chatbot development and API integration. Let's discuss your project for creating a smart search service based on GPT for Bitrix24.

You can review my work cases here: Freelancehunt. I manage projects in various fields, including bot development and CRM system integration. My experience allows me to achieve high conversion rates and efficiency.

I have preliminarily analyzed your task. I am confident that implementing a RAG service and chatbot will significantly simplify working with the Bitrix24 Knowledge Base. What functional requirements are most critical for you? This will help focus on the key aspects of development.

Preparation and launch of the project will take from 2 to 4 weeks, depending on the complexity of integrations and requirements. The estimated cost is from $2000, including development and support.

I suggest discussing the details in private messages or at a meeting. This will allow me to present a more accurate vision of the project and discuss the terms of cooperation.

Am I correct in understanding that your main goal is to create an effective tool for information search and user interaction through the chatbot? Expand

Budget: 600 USD Deadline: 3 days

The task is clear: RAG on top of the Bitrix24 Knowledge Base, a chatbot via imbot.register, and a separate REST endpoint for external services, including your Telegram bot.

I will do the following: set up article import via REST API/Webhooks, chunking and embeddings into a vector database, generating responses strictly based on context with a reference to the source article. I will also set up POST /api/v1/ask to receive questions and chat ID, provide responses, and an array of links, plus Swagger documentation for the bot developer.

Please clarify which vector database is preferred (or at my discretion) and approximately how many articles are in the database. I am ready to show the first working piece (indexing and context-based response) in just 3 days.

Budget: 1200 USD Deadline: 14 days

I already have a practically ready solution for RAG search in the knowledge base and an AI bot, which can be quickly adapted for Bitrix24 and launch the first working version.

Regarding timelines - the target is 10-14 days for the MVP with knowledge base import, re-indexing, GPT responses based on sources, a bot in Bitrix24, and an external API for the Telegram bot.

As for the budget - 600 USD may be enough for a short initial phase with architecture, knowledge base import, and a working search API. I would estimate a full MVP with imbot.register, regular re-indexing, vector database, Swagger or Postman documentation, and protection against responses without sources at around 1200 USD.

I see the implementation as follows - we retrieve articles via REST API or webhooks from Bitrix24, normalize the text, cut it into fragments, build embeddings, save them in a vector database, and GPT only receives the relevant material found and returns a response with links to the articles.

Important questions:
- Is your knowledge base in cloud Bitrix24 or the on-premise version?
- Does the Telegram bot already know how to call external APIs, or does it also need to be developed?

From similar experience:
- https://business.ingello.com/fractal - AI agents and workflow automation
- https://business.ingello.com/vorfahr - AI and SaaS logic with integrations

Our profile for such tasks - https://systems-fl.ingello.com

In touch, we can start with a short technical phase and quickly assemble a working prototype. Expand

Budget: 600 USD Deadline: 3 days

Is there already a webhook for updating Knowledge Base articles, or should reindexing be done according to a cron schedule?

I will import articles via the Bitrix24 REST API, chunking, and embeddings into a vector database, with responses strictly based on context and referencing the source, imbot.register, and the POST /api/v1/ask endpoint with documentation in Postman for the Telegram bot.

I will show the prototype in 3 days.

Budget: 600 USD Deadline: 14 days

Task: raise the RAG service on top of the Bitrix24 Knowledge Base, embed a chatbot within the portal, and provide an external API for Telegram.

I am doing this: a Python service retrieves articles via the Bitrix24 REST API, slices chunks, generates embeddings (OpenAI text-embedding-3-small), and stores them in pgvector or Qdrant. Upon request, we search for top-k chunks, pass them to GPT-4o with the system prompt "respond only based on context," and return the answer + an array of links to the original articles. Simultaneously, I register the imbot via imbot.register, attach an event handler to the webhook. The external POST /api/v1/ask is wrapped in FastAPI, documentation via Swagger, response format: answer / sources[] / status.

Key risk: the Bitrix24 REST API delivers articles page by page and may throttle during reindexing a large catalog, so I set up a queue (Celery + Redis) with rate-limit and an incremental webhook for updating articles.

How many articles are currently in the Knowledge Base, and is multilingual support needed, or is everything in one language?