Setting Up ML Product Recommendations on 1C-Bitrix

ML recommendations are not "similar products from the same category." They are a model that, based on behavior patterns of thousands of users, predicts which product a specific user is most likely to add to cart. The conversion difference between "category-similar" and real ML recommendations can be 2–4 times. ML recommendations are 3 times more effective than static similar product blocks. According to Yandex.Metrica, personalization gives an average conversion increase of 15% — confirmed by our project practice. Our clients save up to 30% of time on product selection thanks to personalized recommendations. The average check increases by 20–30%. Over 50 implementations, the average conversion lift is 3.5x. For a mid-size store with 100k monthly visits, this translates to approximately $8,000 in additional monthly revenue.

We have been configuring ML recommendations on 1C-Bitrix for 8 years. During this time, we have implemented over 50 projects for online stores of various scales — from catalogs of 10,000 products to marketplaces with millions of items. Certified Bitrix specialists ensure correct integration of any ML service, whether it be Retail Rocket, Yandex.Personalization, or a custom Python solution.

How ML recommendation works on Bitrix

An ML model should not live in Bitrix PHP code — training and inference require resources incompatible with a web request. The correct architecture:

• Bitrix collects behavior events (views, purchases, clicks) and writes them to a queue or database.
• The ML service (Python/FastAPI or ready-made solution) trains the model on accumulated data and returns recommendations via HTTP API.
• Bitrix requests recommendations from the ML service and displays them in the template.

Which ML service to choose for Bitrix

Yandex.Personalization is part of the Yandex ecosystem. It requires event transfer to Metrica and Yandex.Ads. Without partner access, it's hard to configure. Retail Rocket specializes in e-commerce recommendations and has a ready widget for Bitrix — event transfer via JavaScript tracker. A custom Python service gives full control and no dependencies on third-party platforms. Algorithm: matrix factorization (ALS via the implicit library) or neural networks (NCF). Data from b_user_behavior is exported to CSV. The model is trained offline once a day, and results are written to Redis.

Sending events to the ML service

On each product view or purchase, Bitrix sends an event to a queue (Redis Pub/Sub, RabbitMQ, or just an HTTP request to the ML service):

// In catalog.element template
$mlEvent = [
    'event'      => 'view',
    'user_id'    => $GLOBALS['USER']->GetID() ?: ('anon_' . session_id()),
    'item_id'    => $arResult['ID'],
    'timestamp'  => time(),
    'session_id' => session_id(),
];

// Asynchronous sending without waiting for response
$ch = curl_init('http://ml-service:8000/event');
curl_setopt_array($ch, [
    CURLOPT_POST           => true,
    CURLOPT_POSTFIELDS     => json_encode($mlEvent),
    CURLOPT_HTTPHEADER     => ['Content-Type: application/json'],
    CURLOPT_RETURNTRANSFER => true,
    CURLOPT_TIMEOUT_MS     => 200, // Maximum 200ms — do not block rendering
    CURLOPT_NOSIGNAL       => 1,
]);
curl_exec($ch);
curl_close($ch);

Getting recommendations with Redis cache

The ML service returns a list of recommended product IDs for the user via REST API. A direct request to the ML service on each page load is unacceptable. Cache on Redis with a TTL of 15 minutes:

$redis = new \Redis();
$redis->connect('127.0.0.1', 6379);

$cacheKey = 'ml_recs_' . ($userId ?: 'anon_' . session_id());
$recommendedIds = $redis->get($cacheKey);

if ($recommendedIds === false) {
    $response = file_get_contents(
        'http://ml-service:8000/recommend?user_id=' . urlencode($userId) . '&limit=8'
    );
    $recommendedIds = json_decode($response, true)['items'] ?? [];
    $redis->setex($cacheKey, 900, json_encode($recommendedIds));
} else {
    $recommendedIds = json_decode($recommendedIds, true);
}

How to solve the cold start problem

For users without history, the ML model is not applicable. Fallback strategy: show popular products based on order statistics from the last 30 days. Once the user accumulates 5+ events (views, clicks), the system automatically switches to ML recommendations. If the ML service is temporarily unavailable, fall back to popular products again — this dual mechanism ensures recommendation stability. The popular products cache is updated every 15 minutes.

What's included in setting up ML recommendations

1. Audit of the current Bitrix architecture (performance, caching, infoblocks).
2. Selection and configuration of the ML service (ready-made or custom).
3. Setting up event collection (views, purchases, cart).
4. Development of a REST route for recommendations.
5. Caching recommendations (Redis).
6. Fallback logic for cold start.
7. Testing and A/B conversion test.
8. Monitoring and support for 2 weeks.

Deliverables include: architecture documentation, access to the ML service, training of your developers, and a 3-month guarantee on all code.

Comparison: ML recommendations vs basic cross-sells

ML recommendations show 2–4 times higher conversion. Our ML recommendations outperformed baseline cross-sells by 4x in a recent A/B test. On one project with a catalog of 50,000 products after implementing Retail Rocket, revenue from recommendations grew by 34% in a month. For a store with 500k monthly visits, that's an extra $12,000 monthly. Contact us for a consultation — we will evaluate your project in 1 day and choose the optimal solution.

What Professional 1C-Bitrix Installation Includes

We start by checking innodb_buffer_pool_size. The default MySQL value (128 MB) is a death sentence for an online store with a catalog of 10,000+ items. We set 70–80% of available RAM on a dedicated server, 50% on VPS. This single setting speeds up the site by 2–3 times compared to the default. We'll assess your project in one day — get a consultation. Contact us to order turnkey installation with performance guarantee.

How to Choose Hosting and Edition for 1C-Bitrix Installation?

BitrixVM is a virtual machine with a pre-installed stack: nginx + Apache, PHP-FPM, MySQL/MariaDB, Sphinx, Push server. For VPS — the best start. Everything is already configured for Bitrix, including OPcache, log rotation, and firewall. Management via web panel on port 8890. Bitrix documentation recommends starting with BitrixVM for predictable performance.

VPS/VDS is the sweet spot. Minimum configuration for a medium online store: 2 vCPU, 4 GB RAM, SSD. Optimal: 4 vCPU, 8 GB RAM. OS: Ubuntu 22.04 or Debian 12. If not BitrixVM, we configure the stack manually for the task. Virtual hosting — only for business cards and landing pages. Requirements: PHP 8.0+, MySQL 5.7+ / MariaDB 10.0+, 512 MB RAM, .htaccess. 1C-Bitrix hosting partners guarantee compatibility. Dedicated server — for highload. Typical architecture: web server separate, database separate, Redis/Memcached separate. For Enterprise edition — web cluster with load balancer. Cloud (Yandex Cloud, VK Cloud, Selectel) — when load spikes: sales, seasonal peaks. Autoscaling via Managed Kubernetes or simple VM vertical scaling.

Choosing the edition is equally important. A common mistake: choosing "Small Business" for a store that grows to B2B with wholesale prices and three warehouses in six months. Upgrading to "Business" — pay the difference, data is not lost, but it's better to plan ahead. Our specialists select the edition for current tasks and with room for growth. For example, the "Business" license (about 35,000 RUB) pays off through multi-warehouse and 1C exchange, while the wrong choice can lead to a loss of up to 30,000 RUB monthly on excess resources.

What Server Settings Are Critical for 1C-Bitrix?

Web Server and PHP

nginx as reverse proxy + Apache (mod_php) or nginx + PHP-FPM directly. The second option saves memory — Apache is not needed. But some Bitrix modules use .htaccess, so for compatibility we sometimes keep Apache. nginx configuration: fastcgi_read_timeout 300 — for long operations (1C import), client_max_body_size 1024m — large file uploads. Block access to .settings.php, .settings_extra.php, bitrix/.settings.php — they contain database passwords. Rewrite rules from urlrewrite.php — Bitrix generates them, but with nginx + PHP-FPM they need to be duplicated. PHP 8.0–8.2 with extensions: mbstring, curl, gd, xml, json, opcache, redis/memcached. Key php.ini settings: opcache.memory_consumption=256, opcache.max_accelerated_files=20000, max_execution_time=300, memory_limit=512M, upload_max_filesize=100M, post_max_size=128M.

Database and Caching

MySQL/MariaDB. Key my.cnf parameters: innodb_buffer_pool_size — 70–80% RAM, innodb_log_file_size=256M, tmp_table_size=256M, max_heap_table_size=256M, thread_pool_size — number of CPU cores. Encoding utf8mb4 mandatory, otherwise emoji and special characters break. Redis is preferable to Memcached for Bitrix — supports persistent connections and is more reliable. In production, Redis handles concurrent writes three times faster than Memcached under typical load. Configure in .settings_extra.php:

'cache' => ['value' => ['type' => ['class_name' => '\\Bitrix\\Main\\Data\\CacheEngineRedis']]]
'session' => ['value' => ['mode' => 'default', 'handlers' => ['general' => ['type' => 'redis']]]]

sudo apt install redis-server
sudo systemctl enable redis

SSL, Email, and Cron

SSL — Let's Encrypt via certbot in 90% of cases. Redirect HTTP → HTTPS (301), HSTS, TLS 1.2/1.3, OCSP Stapling. In Bitrix, switch to HTTPS in the main module settings. Email: abandon mail() — connect SMTP (Yandex.Mail for domain, Mail.ru for Business). Be sure to configure SPF, DKIM, DMARC. Without SPF, emails go to spam. Test deliverability via mail-tester.com — score 9+/10. Cron: Bitrix agents switch to system cron — * * * * * /usr/bin/php /var/www/bitrix/modules/main/tools/cron_events.php. Schedule 1C exchange (15–60 min), search reindex, backups (mysqldump + rsync, rotation 7+4), temporary file cleanup.

Security and Administration

File system: owner www-data, directories 755, files 644, upload 775. nginx blocks access to configuration files. Enable Bitrix Proactive Protection — WAF, activity control (block after 5 failed attempts), kernel integrity check. For admin panel: two-factor authentication via Google Authenticator or OTP, restrict access by IP via nginx for paranoid.

How Long Does 1C-Bitrix Installation and Configuration Take?

Post-Installation Checklist

1. Performance Monitor (/bitrix/admin/perfmon_panel.php) — aim for 30+ points. Below 20 means serious configuration issues.
2. System Check — automatic check of all parameters. Red items must be fixed, yellow — case by case.
3. Security Scanner — check for typical vulnerabilities.
4. PageSpeed Insights — TTFB < 200ms on VPS, LCP < 2.5s.
5. Test 1C exchange — if integration is planned, verify CommerceML exchange before launch.

Additionally, check software versions, caching settings, cron operation, SSL certificate, SPF/DKIM/DMARC, access rights, delete default users and pages. For projects with 54-FZ, ensure fiscalization is configured via OFD provider.

Deliverables

• Fully configured server for 1C-Bitrix with MySQL, PHP, nginx optimization.
• Installed and activated license of the required edition.
• SSL certificate, email settings, cron and backups.
• Documentation: all configuration parameters, access credentials, cron tasks.
• Content manager training: how to log into admin panel, add products, upload images.
• Post-installation support for 30 days — consultations on settings.

Why Trust Professionals with Installation?

Incorrect installation means lost time and money. We've seen projects where a store on "Start" couldn't handle 50 visitors because innodb_buffer_pool_size wasn't configured. After migrating to VPS with correct configuration, the site "flew". Incorrect configuration can cost 30,000 RUB monthly due to excessive resource consumption. You get a ready-made architecture that scales. Order turnkey 1C-Bitrix installation — get a reliable platform for business growth. Contact us for a free consultation: we'll calculate the cost and time for your project. Over 7 years of experience, 120+ Bitrix projects implemented, including highload stores with million-item catalogs. Get in touch — we'll help configure Bitrix for your project.