PredictST Consulting

Forecast what happens
when and where.

I help companies extract value from location and time-based data. PhD-backed spatio-temporal prediction to forecast demand, optimize resources, and reveal hidden patterns.

Let's Talk See Use Cases

What I Do

Capabilities

Demand & Load Forecasting

Predict usage, traffic, or resource demand at specific locations and times. Whether it's network load, delivery volume, or foot traffic — know what's coming before it arrives.

Pattern Discovery

Find clusters of similar behavior in your spatial data. Identify locations, routes, or time periods that behave alike — insights that aren't visible to the naked eye.

Anomaly Detection

Spot unusual patterns that deviate from the norm. Applications range from fraud detection to network monitoring, quality control, and infrastructure safety.

Custom Model Development

Have a unique spatial-temporal challenge? I build tailored prediction models using state-of-the-art Deep Learning methods designed for your specific data and business needs.

Applications

Where Spatio-Temporal AI Applies

These methodologies are sector-agnostic. If your data has coordinates and timestamps, it can be modeled.

Fig 1.1: Network Topology Match

Traffic & Transportation

Predicting traffic speeds, travel times, and congestion patterns. My research demonstrated how to forecast traffic in cities with no historical data by leveraging road network topology and transfer learning.

Network Topology Transfer
ETA Prediction
Logistics Route Optimization

Fig 2.0: Spatial Clustering

Telecommunications

Forecasting cell tower utilization, predicting network load patterns, and identifying clusters of similarly-behaving infrastructure. Turn your spatial network data into operational foresight to prevent outages.

Load Balancing
Infrastructure Planning
Anomaly Detection

And Beyond

The same mathematical foundations apply across industries. If you have moving assets or location-dependent demand, let's talk.

Energy

Grid load forecasting & consumption

Retail

Site selection & foot traffic

Ph.D. Spatio-Temporal Prediction
L3S Research Center Alumni

About Me

I predict things that move through space and time.

My journey started at Hannover Re, where I built prediction models for soccer player injuries and automated quotation tools. That early exposure to risk and probability led me to a PhD at the L3S Research Center, one of Germany's top AI institutes.

There, I spent six years developing novel methods for traffic forecasting, trajectory analysis, and transfer learning. I focused on the hard problems: How do we predict traffic in a city where we have no sensors? How do we find patterns in chaotic data?

I've led work packages in major EU research projects and collaborated with industry leaders like Volkswagen AG on real-world mobility challenges. Now, I operate as an independent consultant, helping forward-thinking companies turn their location and time data into actionable foresight.