Technology partners

Who we build with

Our partners

Influx Data
Hypernode

In more detail

What we use them for

Orq.ai
A platform for building and running LLM features: prompts, evaluations and monitoring in one place.
Why Orq.ai
- Prompts and model settings live in one place, instead of being scattered through the code.
- We can compare models and prompt versions before a change goes live.
- It is part of the stack behind the AI assistant we run for Victron Energy.
- We compared it with LangChain, LangSmith and LangGraph in this article.
BigCommerce
A hosted eCommerce platform for brands that outgrow standard webshop software.
Why BigCommerce
- Hosted for you: no servers to maintain, and updates are handled by BigCommerce.
- Handles large catalogues with lots of products and variants.
- Open APIs, so it connects to your ERP, PIM or a custom frontend.
Influx Data
The company behind InfluxDB, a database made for time-series data like sensor readings and energy measurements.
Why Influx Data
- Made for measurements: data points from sensors, inverters and batteries, stored and queried by time.
- For Victron Energy's VRM platform we worked together with Ytec and Influx to make the system fast and reliable for everyone who monitors their energy system.
- Dynamic ESS uses InfluxDB for the historical data it needs to plan the most cost-efficient battery schedule.
Hypernode
Managed hosting built specifically for Magento webshops.
Why Hypernode
- Hosting made specifically for Magento webshops.
- We are a Hypernode partner through Byte, so we arrange and manage the hosting for you.
- Fixed monthly hosting costs, so there are no surprises later on, like for the FrieslandCampina webshop.
- The official Segway-Ninebot webshop for Europe runs on it too.
Vitess
A way to scale MySQL across many servers. GitHub and Slack use it for their largest databases.
Why Vitess
- It speaks MySQL, so your application doesn't need to change.
- The database is split over several servers (sharding) without you having to build that into your code.
- If a server fails it recovers by itself, and data migrations run with almost no downtime.
- We helped Victron Energy move very large databases to Vitess.
ClickHouse
A database built for analytics: questions over billions of rows, answered in a fraction of a second.
Why ClickHouse
- Fast enough for live dashboards on top of very large amounts of data.
- Grows with you: add servers when the data grows, and it keeps taking in new data while you query it.
- We use it to replace slow data warehouses and reporting backends, for dashboards, product analytics and monitoring.
Neo4j
A graph database: made for data where the connections matter as much as the data itself.
Why Neo4j
- Questions like "who worked on what, with whom, for which client" stay fast, even with millions of connections.
- Fits domains that are hard to squeeze into tables, like knowledge graphs, supply chains and recommendations.
- Our own OKREBOT uses a knowledge graph in Neo4j to know which people, projects, clients and decisions are connected.