Customers
How our customers work with zwrm.
An online real estate platform, a stock market newsroom and a technology company: at all three, agents work inside the tools the team already uses. Here’s how they got started and what has changed since.
- Customer reply drafted
- Invoice corrected in the ERP
- Bug filed as an issue
- Daily summary posted
At Argetra, anyone can now hand off work that used to take hours.
Argetra runs an online real estate platform where prospective buyers find foreclosure auctions across Germany. Its operations team connected zwrm agents to the company’s tools, one for each task. Since then, anyone on the team can hand off work.
The starting point
Running the platform involves a lot of recurring work: requests in the inbox, entries in the ERP system, tickets, coordination within the team. Some of these tasks took hours every time. Often, only someone who knew every system could do them.
The approach
The agents are connected to Microsoft 365 with Outlook and Teams, plus Notion, Linear and Cloudflare. In-house systems such as the ERP connect via MCP. Each task has its own agent, and each agent only has the access it needs for that task. To hand something off, people message the right agent in Teams or forward it an email. The agent does the work in the tools where it comes up and checks in when it needs approval.
The result
Tasks that used to take hours are now handled by the agents. Anyone at Argetra can hand them off, not just the operations team that set the agents up. The people who know the systems best have time again for work on the platform itself.
- 7
- Integrations connected
Where the agents work
- Microsoft 365
- Outlook
- Microsoft Teams
- Notion
- Linear
- Cloudflare
- MCP
“zwrm agents are now integrated into our email, Teams, and ERP system. Everyone can just delegate tasks to zwrm that used to take hours on end.”
- Prices updated in the app
- Article image created
- Posts scheduled for the week
- Weekly reach reported
Börse Inside builds its own apps, and zwrm keeps them running.
Börse Inside covers the stock market, in investment newsletters and on social media among other channels. Its editors build their own apps on zwrm. Because zwrm takes care of hosting and maintenance, the apps keep running.
The starting point
A stock market newsroom works with a lot of data every day, from share prices to social media reach. Off-the-shelf apps rarely fit, and the newsroom doesn’t have a development team of its own. And every app you build yourself needs someone to run and maintain it.
The approach
Editors tell an agent what an app should do, and the agent builds it on zwrm. The agents are connected to 12 data sources and know which data belongs to which app. They pull figures from PostgreSQL, create images and videos with Higgsfield and schedule posts in Metricool. They talk to the newsroom in Teams and file documents in Microsoft 365.
The result
The newsroom now builds apps it would once have needed developers for. Nobody in the newsroom has to deal with hosting or maintenance, and the apps keep running. Good ideas no longer have to wait until someone has time to code them.
- 12
- Data sources connected
Where the agents work
- Microsoft 365
- Microsoft Teams
- PostgreSQL
- Higgsfield
- Metricool
“zwrm enables our whole editorial staff to build apps that stick around. We don’t have to worry about maintenance or hosting because the AI takes care of it for us. We’ve connected 12 different data sources to our agents, so they always know which data belongs to which app. It’s like having a 10-person developer team.”
- Customer request assigned
- Receipts collected for the tax advisor
- Follow-up emails scheduled
- Error report analysed
At Elan Technology, the agents work in the legacy systems too.
Elan Technology built its own internal agent platform on zwrm. It’s connected to the company’s tools and to its specialised legacy systems as well. It’s now part of everyday work.
The starting point
Alongside common tools such as Front, Slack, Notion and DATEV, Elan Technology works with specialised legacy systems. They hold a lot of data and processes, but they weren’t built for AI. So IT went looking for a platform that gives agents access to all of these systems and that anyone in the company can use.
The approach
The platform’s agents work in Front, Notion, Slack, lemlist, DATEV, GitHub, PostHog and Sentry. The legacy systems are connected too. Anyone who needs something asks an agent in a chat. Automations and routines handle recurring work on their own. That way, customer service, sales, accounting and engineering all work on one shared platform.
The result
What started as an internal project is now used across the company: more than 500 chats a day run on the platform, plus dozens of automations and routines. Because the legacy systems are connected, the agents can also reach data that only lives there. New workflows go onto the same platform instead of into yet another tool.
- 500+
- Chats a day
Where the agents work
- Front
- Notion
- Slack
- lemlist
- DATEV
- GitHub
- PostHog
- Sentry
“We built our own internal agent platform with zwrm. zwrm was the only platform that let us connect our specialised legacy systems to AI. Our colleagues now start more than 500 chats a day, on top of dozens of automations and routines.”
Tom Jordi RueschChief Information Officer, Elan Technology Tell us what lands on your desk every week. We’ll show you an agent doing it.
A demo with our team, in English or German. Bring one recurring task and the tools it touches.