Overview
Founded in Hasselt by Bert Baeck, TrendMiner developed self-service analytics software for process and manufacturing industries. The platform enabled process engineers and plant operators to search, detect, and predict patterns in time-series sensor data from industrial plants, without requiring data science expertise.
TrendMiner's customer base included major industrial companies such as Pfizer, BASF, and Total, demonstrating the platform's applicability across chemicals, pharmaceuticals, and energy. The company was acquired by Software AG, integrating its process analytics capabilities into the German software giant's industrial IoT portfolio.
The acquisition represents one of Belgium's notable industrial software exits and validated the market for self-service operational analytics in capital-intensive industries. TrendMiner shared the Hasselt base with fellow manufacturing-focused startup Arkite.
Key people
In the news
- Process manufacturers spend more time chasing data than acting on it. That's what we'll be talking about at SPS 2026. TrendMiner heads to Nuremberg November 24–26, sharing a booth with the Open Industry 4.0 Alliance — Hall 5, Stand 160. Two teams, one conversation: what interoperable, self-service analytics actually looks like on a real plant floor. If root-cause analysis, predictive insights, or plant-wide optimization are on your list for 2027 planning, put fifteen minutes on our calendar in Nuremberg. Book a slot:
- At a specialty fiber plant, nine-hour batches kept running longer, and conventional tools couldn't explain why. The answer was already in five years of historian data. The engineers filtered out every batch that didn't count and layered the shortest ones into a golden batch profile. When they laid ten recent batches on top, the gap showed up midway through the cycle: a raw material feed was ramping up too slowly. They found the root cause in four hours instead of days. The improvement potential: an estimated $1M a year. See how
- TrendMiner 2026.R3 puts an AI agent inside the application, right next to the trend you’re investigating. The TrendMiner Agent knows which view you’re on. It searches tags, runs value-based searches, adds tags to TrendHub, changes the time range, and answers “how do I” questions in context. It runs on your own LLM provider, so your data stays inside your company landscape. The release also adds three MCP tools for context items and monitor triggers, logarithmic scale in TrendHub, and a Query Designer for ODBC and JDBC sources. Read
- Avebe’s protein coagulator tripped 44 times in three months, and the root cause was hard to find. The process experts didn’t wait for a data science project. They went into their own historian data. Similarity search pulled up every trip at once. Layer comparison put the bad periods next to the good ones, and one signal stood out: process water flow dropped whenever pressure built up in an upstream flash tank. In just four hours, they reduced the tripping by 50%. Read the full Avebe story: https://hubs.la/Q04BSthc0
- The TrendLab Europe 2026 aftermovie is out. Two days at the Evoluon in Eindhoven, condensed into a few minutes. Most of the stage time went to customers rather than to us. Process experts walking through problems from their own plants and the analytics that solved them. The film shows the room. The slides show the method, and the customer presentations from both Eindhoven and Houston are up in the TrendMiner Community website. If you were there, what is the one thing you took home and actually put to work? Europe slides:
- Trade show calendars fill up fast, and the good 1:1 slots go first. With five weeks until SPS 2026, we're already booking meetings at our Hall 5, Stand 160 booth — shared with the Open Industry 4.0 Alliance. If your plant is still running root-cause analysis and yield troubleshooting manually, this is worth fifteen minutes to see how Industrial AI & Analytics changes that math. Grab a slot before the calendar closes: https://hubs.la/Q04yz6Y60
- Monument Chemical didn’t need new data to cut time to a shippable tank by about 60%. They needed a faster way to read what was already in their historian. When a customer asked for higher purity, their Houston engineers went back through years of production runs, compared them, and found a way to deliver it without adding much production time. The question came from the customer. The answer came from their own process history. Monument Chemical wrote this one up themselves, so it’s worth reading in their words. Thanks to Vincent
- Monument Chemical cut the time to a shippable tank after catalyst activation by about 60%, using data they already had. At TrendLab 2026 North America, Yndalecia Martinez showed how the Monument Chemical Houston team uses TrendMiner to compare production runs, spot patterns, and act faster. When a customer needed a higher-purity product, the engineers went back through years of production data and found a way to deliver it without adding much production time. Automated reporting and integrated environmental monitoring are also
Open roles
- Unsolicited ApplicationBelgium,Spain,Germany
Alumni 1 went on to found or lead
Bert BaeckCEO→CEO at Timeseer.AI
Related profiles
Something wrong or missing? Send an update. Fixed within 24 hours.