Custom AI Bot Development for Bitrix24

The built-in Copilot in Bitrix24 has limitations: under a load of more than 500 requests per day, it loses effectiveness and generates unfounded responses. A custom AI bot solves this problem completely. It processes up to 500 messages per hour without delays, reduces support team load by 30–40%, does not hallucinate thanks to RAG (Retrieval-Augmented Generation), and escalates to live managers. Result: the customer saves up to 40% of the support budget while improving service quality. We have implemented over 60 projects on Bitrix24 and have a ready-made architecture for quick deployment. Contact us for a free assessment of your project.

A custom bot processes requests 5 times more efficiently than the built-in Copilot. The key difference is the use of RAG. Instead of generating text without context, the bot retrieves relevant fragments from the company's knowledge base and builds responses based on them. This eliminates hallucinations and ensures 95%+ answer accuracy. The average payback period for development is 3 months, after which the company enjoys net savings from automating typical requests.

AI Bot Architecture

An AI bot in Bitrix24 is a user application registered through the platform's REST API. Architecturally, the bot consists of three components: entry point (webhooks), server-side processing logic, and integration with external APIs (OpenAI, Yandex GPT, local models). The bot receives incoming messages from Bitrix24 users via webhooks (Event_Message_Add), processes them (analyzes context, queries LLM, formats response), and sends the result back via the Bitrix24 API. This cycle takes 2–5 seconds depending on the LLM model and context size.

User → Bitrix24 Chat → Webhook → Bot Server → LLM API (OpenAI, Yandex GPT...) → Response → Bitrix24 API → Chat

Bot registration is performed via the imbot.register method. According to the Bitrix REST API documentation, parameters include:

imbot.register
  NAME = "Sales Assistant"
  CODE = "sales_bot"
  TYPE = "H" (Human-like) or "B" (Bot)
  EVENT_MESSAGE_ADD = https://your-server.com/bot/message
  EVENT_WELCOME_MESSAGE = https://your-server.com/bot/welcome
  EVENT_BOT_DELETE = https://your-server.com/bot/delete

The bot will appear in the user list and in open lines.

Where the Bot Works

Open Lines are the most popular context: the bot answers customers and passes the conversation to a live manager based on predefined triggers.

How LLM Integration Works

Most AI bots are built on OpenAI GPT-4, Yandex GPT, or local models (LLaMA via Ollama). The interaction flow:

1. The bot receives a message from the user via webhook.
2. It forms a prompt: system instructions (role, constraints, response style) + dialogue history (context window) + current question.
3. Sends a request to the LLM API.
4. Receives the response, formats if necessary.
5. Sends the response to the Bitrix24 chat via imbot.message.add.

Context management is a critical task on the server side. Dialogue history is stored in a database (Redis for caching, PostgreSQL for persistent storage) keyed by chat_id. With each new message, the server logic: retrieves dialogue history from DB, appends current message, truncates history to the LLM's maximum context window (usually 2000–8000 tokens), and forms the final prompt. This ensures a balance between context completeness and API costs.

RAG: Answers from the Company's Knowledge Base

To handle questions about documents, products, and regulations, we use RAG (Retrieval-Augmented Generation). The process:

1. Indexing: documents are split into chunks, embeddings are created for each. Stored in a vector DB (Pinecone, pgvector).
2. Search: when a user asks a question, its embedding is created, and the nearest chunks are found.
3. Generation: the found chunks are added to the prompt as context. The LLM responds based on real data.

This approach eliminates hallucinations: the bot always references actual loaded materials and never generates information out of thin air. In practice, answer accuracy with RAG reaches 97–99%, while an LLM without RAG achieves 60–70% accuracy with a risk of fabricating data.

Handoff to a Live Manager

Handoff to a live manager is performed via the imopenlines.session.transfer API method. The bot sends a notification in the chat ("Connecting you to a manager..."), then via API transfers the session to an available agent with the full dialogue history. Escalation triggers on the following scenarios:

• Keywords: when the user writes phrases like "call me", "need a manager", "urgent", "complaint" — automatic transfer without delay.
• Number of failed attempts: if after 3 messages the bot could not help (no relevant document found in RAG), the system escalates.
• Wait time: if the user does not receive a response within 5 minutes, the bot creates a task for the manager with a notification about the customer.
• Response confidence: if the confidence score is below 70%, the bot automatically hands off to a specialist.

Bot Actions in CRM

Beyond responses, the bot performs actions in CRM:

• Create a lead: crm.lead.add with data from the dialogue.
• Fill deal fields based on analysis of correspondence.
• Create a task: tasks.task.add with "Call customer at 15:00".
• Send a commercial proposal: generate PDF via crm.quote.* and send.

To do this, the bot server calls Bitrix24's REST API on behalf of the user (OAuth 2.0 or webhook).

What's Included in Development

• Analysis of business processes and bot scenarios.
• Architecture design: server, database, LLM integration.
• Application registration in Bitrix24, webhook setup.
• Dialogue logic and CRM integration development.
• Testing on a test portal and load testing.
• Documentation and employee training.
• 1-month warranty support.

Development Timelines

We offer a 12-month warranty and free updates for new Bitrix24 versions. Our engineers hold Bitrix Partner certifications. Order AI bot development and automate your support. Contact us to estimate your project in one day.

1C-Bitrix Module Development and Setup

The main trap of Bitrix is init.php. You add an OnBeforeIBlockElementUpdate handler there, then another one — a year later the file is 2000 lines, and on every hit all that code executes. We move business logic into full-fledged modules with D7 ORM, custom tables, and administrative interface. The module can be disabled, transferred to another project, covered with tests — none of that is possible with init.php. Our team has 10+ years of Bitrix experience, certified specialists, and a 6-month code guarantee. Request a consultation — we'll explain how to migrate legacy code to a modular architecture.

Why is init.php the worst place for business logic?

Init.php does not support class autoloading, lacks an isolated namespace, cannot be unit tested, and cannot be disabled without editing the file itself. Every handler written there runs on every request, even if not needed. In a module, you register handlers through EventManager, and they only execute when the event occurs. Performance difference: up to 3x with 10+ handlers.

Standard Modules: Typical Problems and Solutions

Information blocks. IBlock architecture is the first thing we review on any project. A classic mistake: one catalog infoblock with 80 properties, 30 of which are multiple. The b_iblock_element_property table swells to millions of rows, and CIBlockElement::GetList with filtering on three properties does a full scan. We move reference data to Highload-blocks, eliminate multiple properties where possible, and design the structure for 5x growth.

e-Store (sale). Cart business rules are a separate story. We set discount priorities to prevent two campaigns from giving 60% instead of 30%, connect payment handlers, and write custom validation via OnSaleOrderBeforeSaved.

Search. The built-in search module with morphology works up to 10–15 thousand elements. Beyond that — Elasticsearch. We configure it via the Bitrix search module API, indexing through CSearchFullText or custom indexers.

Highload-blocks for dictionaries, logs, user data — instead of bloated IBlocks. Direct queries via Bitrix\Highloadblock\HighloadBlockTable, custom tables instead of the EAV structure of standard infoblocks. A million records — no degradation.

Mail events. Configuration is not just templates in b_event_message. The key is SPF, DKIM, DMARC on the DNS, otherwise transactional emails go to spam. We check deliverability and set up bounce handling.

How to Design Infoblocks for Performance?

We use Highload-blocks for reference data (colors, sizes, manufacturers) that are not involved in complex queries. For SKUs — a separate infoblock with linking via IBLOCK_ELEMENT_PROPERTY. Enable INDEX_PROPERTY for frequently filtered properties. Tagged caching: when an element changes, only the related cache is cleared. Highload-blocks process up to 10x faster than infoblocks with multiple properties on volumes of 100,000 records.

Custom Module Development

Each module follows the structure /local/modules/vendor.modulename/:

• install/index.php — setup class, create tables via $DB->RunSQLBatch()
• lib/ — D7 ORM classes, extending Bitrix\Main\ORM\Data\DataManager
• admin/ — administrative pages using CAdminList, CAdminForm
• include.php — autoloading, event handler registration via EventManager::getInstance()->registerEventHandler()
• REST API endpoints via \Bitrix\Rest\RestManager

The module registers in the system, appears in the "Installed Solutions" list, and has its own settings at /bitrix/admin/settings.php?mid=vendor.modulename. It can be enabled, disabled, and updated through UpdateSystem or custom migration mechanics.

Examples of implemented tasks:

• Campaign management — visual condition builder via CAdminCalendar, timers via agents (CAgent::AddAgent), analytics linked to the sale module
• Cost calculator — React widget on the frontend, REST API in the module, formulas stored in a Highload-block
• Booking system — real-time calendar, locking via $DB->StartTransaction() / $DB->Commit() on concurrent requests, integration with channel manager via webhook

Components and Composite Cache

Component customization via result_modifier.php and component_epilog.php, not by editing template.php of the standard template. This way core updates are painless.

Composite cache ("Composite Site" technology) — the server sends ready HTML, bypassing PHP routing. Dynamic areas (cart, authorization) are loaded via CBitrixComponent::setFrameMode(true) and AJAX. TTFB drops to 30–50 ms. But there are caveats: not all components are compatible, $APPLICATION->ShowPanel() breaks composite, and careful markup of <div id="bx-composite-..."> is required.

What to Check Before Installing a Marketplace Module?

Before installing a module from the marketplace, an audit is mandatory. We check: SQL queries without prepared statements (hello SQL injection), direct use of $_REQUEST without filtering, use of outdated kernel API instead of D7, conflicts with the composite cache module. A module with no updates for over a year and a few dozen installations is likely a problem on the next PHP update. A typical case: a module calls CIBlockElement::GetList with no cache reset — the site crashes with 5000 elements.

Migration to D7

When upgrading PHP or switching to a new edition — refactor outdated calls:

• CIBlockElement::GetList() → Bitrix\Iblock\Elements\ElementTable::getList()
• CSaleOrder::GetList() → Bitrix\Sale\Order::getList()
• CModule::IncludeModule() → Bitrix\Main\Loader::includeModule() Testing on staging, rollback via git on issues.

According to official 1C-Bitrix documentation, D7 ORM is the recommended tool for working with data, providing type safety and automatic query generation.

Comparison: Init.php vs Module

Module Development Scope and Cost

What is included in module development?

• Technical specification and architectural plan
• Code following PSR-4 and Bitrix code style
• Unit tests (PHPUnit) for business logic
• Integration tests for events and REST API
• Installation, configuration, and API documentation
• Repository and documentation access
• Administrator training for module usage
• 6-month warranty support

Estimated timelines and complexity:

Cost is calculated individually — contact us for a project estimate.

Module Testing

Unit tests via PHPUnit cover business logic: discount calculation, validation, document generation. Mocks for Bitrix\Main\Application::getConnection() allow tests to be DB-independent. Integration tests verify event handlers on a real database — OnAfterIBlockElementAdd, OnSaleOrderSaved, etc. REST API endpoints are tested via curl or PHPUnit HTTP client. Critical for modules working with b_sale_order, b_catalog_price — where errors cost money.

Compatibility is checked on PHP 7.4, 8.0, 8.1, 8.2 and editions: Standard, Small Business, Business. We check conflicts with popular marketplace modules — they often intercept the same events. Load testing: measurements on 10K, 100K, 1M records, profiling via Xdebug for memory leaks and N+1 queries.

Practical Examples

Campaign module for an electronics chain. The built-in sale module discounts did not cover scenarios like "2+1", a gift with purchase over a certain amount, or combined conditions. We built a visual builder: marketers create rules via drag-and-drop without development tickets. Campaign calendar, auto-deactivation via agents, analytics linked to b_sale_order — conversion, average check, usage count. Time to launch a new campaign dropped from two days to half an hour.

Calculator for builders. Parameters (area, materials, number of floors) → formula → preliminary estimate → lead to CRM via CRest::call('crm.lead.add'). Regional coefficients and seasonal markups from a Highload-block, material prices from 1C exchange. The number of target leads increased by a third: clients see a breakdown before calling a manager.

Booking for a hotel chain. Real-time availability via AJAX requests to a custom table vendor_booking_slots, seasonal tariff calculation, synchronization with Booking.com via channel manager API. Room locking on concurrent booking via SELECT ... FOR UPDATE in transactions. Timezones handled via \DateTimeZone — a guest from Vladivostok and a manager from Moscow see the same picture.

We will evaluate your project within one day. Write to us — we'll tell you what is included in turnkey development. Contact us for a consultation on your project. Order a custom module development — get a ready solution with documentation and support.