Reservix turned its CMS into an AI-driven sales engine

Managing content for a large data-heavy platform can be very time-consuming.

Reservix wanted to help its content admins get things done faster, but also use the opportunity to turn its CMS into a tool that contributes more to the sales effort.

Specific goals included:

• reducing manual effort to manage content,

• leveraging AI to speed up content creation,

• increasing the visibility of events across multiple publishing channels.

To that end, Reservix partnered up with The Software House.

The client

Reservix is Germany's leading event ticketing platform, operating across the German-speaking market under two big brands, Reservix and ADticket.

The company serves audiences in Germany and Austria across music, comedy, theatre, sports events, and live entertainment.

Each brand has its own audience and content strategy, enabling event organizers to reach a variety of specific audiences.

INDUSTRY

Ticketing and live entertainment

COUNTRY

Germany

SERVICE PROVIDED

Full-stack development, cloud architecture, AI and ML integration, DevOps

Business context

The Reservix admins managed a large volume of event-related content through a dedicated content management system.

The company sought ways to reduce the amount of manual labor required daily and to leverage the large volume of data to power sales efforts.

The content managers wanted a flexible CMS that would let them:

• group events into marketable containers,

• publish content across multiple brands through a single workflow,

• feed data to external sales channels to reach a wider audience.

Achieving all that required new architecture, integrations, and a Node.js development partner to join Reservix’s in-house team.

Challenges

Here are some of the specific requests that needed to be addressed:

1. Content managers needed a structured way to group and market events across Reservix, ADticket, and other brands,

2. The team aimed to unify publishing across Reservix and ADticket under a single content workflow,

3. Event data needed to reach external sales platforms such as Spotify and Google,

4. The team wanted AI tools that could help managers match audiences with events likely to interest them.

5. Updates for thousands of events needed efficient processing as the platform scaled,

6. External integrations required real-time data synchronization across all connected systems.

The new and modernized CMS would feature time-saving AI functionalities

Solution

Team formation

Reservix and The Software House formed a joint team that combined Reservix's product and ecosystem expertise with TSH's full-stack and AI engineering.

Technology choice

• Node.js,

• React,

• PostgreSQL,

• GraphQL,

• Express,

• AWS Lambda, SQS, SNS, Step Functions, S3,

• Serverless architecture,

• Python (AI modules),

• Amazon Bedrock (LLM communication on AWS),

• Algolia (AI-driven search and recommendation service).

Process

The work progressed across 4 stages, each building on the previous one.

Stage 1: Branded Containers

The first stage gave Reservix content managers a new way to organize and publish events across different brands.

Reservix and TSH jointly designed the Branded Container model, a persistent marketing entity that groups related events under a single object.

A container could include an artist, a genre, a venue, a tour, or an event series.

New possibilities unlocked

Containers persist over time, so a manager can market an artist or a venue across all their upcoming events from a single object rather than on an event-by-event basis.

Data containerization streamlines drafting and variant creation, letting you publish independently across channels and brands instantly.

Stage 2: External channel integrations

Once the core platform features were delivered and had become stable and scalable, the teams connected the platform to 4 external systems.

The goal was to support a multi-channel strategy by automating marketing and expanding reach, boosting the brand's visibility and accessibility wherever current and potential customers are active.

Mailon

The Content App connected to Mailon, the newsletter and fan mailing system, so that mailing logic was driven by container data managed inside the platform.

As a result, the Content App could send dynamic content not only to the ticketing service but also to dynamically generated newsletters.

Spotify

Integration with Spotify was the next step to build multichannel sales.

Event data was pushed to Spotify so that users searching for artists could see upcoming events and navigate directly to the Reservix ticket shop.

It created new, highly targeted sales opportunities and helped build brand awareness among Spotify users, noticeably increasing the rate at which Spotify recipients turned into buyers.

Google

Next was integration with Google that required a dedicated solution to send the event and artist data to Google Actions Center in cyclical batches.

The priority here was to optimize the process of sending big chunks of data. So Reservix architects and the TSH team jointly defined which event types met Google's acceptance criteria.

The branded container included the artist or venue relationships required by external platforms, providing a clean way to expose relational event data.

Stage 3: AI-generated content and suggestions

With the container model running, the Reservix and TSH teams brought AI into the content team's daily workflow.

With AI, the joint team could turn the Content App into a real-time data source for Reservix's wider sales and marketing ecosystem.

Container-event assignment

The LLM analyzed incoming events, compared them against existing containers, and surfaced the closest match.

Managers received ranked suggestions rather than browsing a full event catalog, reducing the time spent on manual review and making the whole process more efficient.

New container suggestion

When a new container made sense, the system proposed creating one.

Then, it generated draft descriptions for Reservix and ADticket from event data, keeping a human in the loop to review and refine both brand outputs simultaneously.

Suggestion engine

A filter-based suggestion engine proposed which events to assign to which containers, giving managers a curated shortlist to work from.

Algolia

The platform also integrated with Algolia, replacing OpenSearch with a dedicated search service.

Algolia enabled AI-ranked results, dynamic data ranking, and real-time index updates driven by container and event changes.

Changes to containers or events trigger immediate Algolia updates through an event-driven notification pipeline, with no polling required.

The key challenge was updating a high volume of data on the fly.

Thanks to the implementation of Algolia, data stored in the Content App and marked as ready for display is instantly available in the search engine, ensuring web shop end users always see up-to-date information.

Stage 4: Architecture and performance

As data volumes and integration demands grew, the joint team focused on making the platform reliable and efficient at scale.

Batch processing

To speed up and scale up all the aforementioned integrations, the teams moved to a batch-processing model, grouping updates before processing, applying this pattern across the whole data pipeline as load requirements grew.

The Algolia integration was the first to use batch processing by default.

Branded containers make it easier to use a group of related events for marketing purposes like targeting segmented audiences

Outcome