Engineering, understood.AI, applied.
We advise, train teams, work inside development projects and supply the software. By engineers, for engineers, based in Germany and Austria and working with clients across Europe and beyond.
Simulate for hours. Or predict in milliseconds.
This is what a surrogate model feels like: change a process parameter and see at once how the wall thickness of the part responds. No new simulation.
Move a slider, watch the prediction.
- Minimum
- 1.47 mm
- Mean
- 1.94 mm
- Maximum
- 2.36 mm
Illustrative model to play with. Parameter ranges follow NAFEMS Magazin 1/2024; not project data.
Documented results from published projects
Four services, one partner
Start wherever you are. Each service works on its own and connects to the others.
Find use cases, check your data, build the roadmap and business case.
Executives, CTOs, heads of development, simulation and testing→ Learn moreSeminars, workshops and coaching on your first own project.
Simulation, development, test and CAE teams, project leads→ Learn moreEngineering and R&D projects, pilot projects and proof of value.
Heads of R&D, project leads, technical procurement→ Learn moredAIve Predict in the browser or on premise, for individuals and teams.
Engineers, specialist teams, IT and procurement→ Learn moreWhere are you right now?
Featured project
Predicting residual wall thickness in water-assisted injection molding
A neural network predicts the residual wall thickness of a hollow plastic part from its process parameters. Run in reverse, it returns the process setting needed for a target wall thickness. A project with Plastic Innovation.
- Industry
- Plastics – water-assisted injection molding (WIT)
- Partner
- Plastic Innovation GmbH
- Method
- Process simulation, MLP (3 × 32 neurons), forward and inverse prediction
- Data
- Simulation data: 4 process parameters, wall thickness at 300 positions
- Software
- dAIve Predict (no-code ML)
- Published
- NAFEMS Magazin 1/2024
Our method
Physics and data belong together. That is how models become fast and still reliable.
Define the problem
Build the model, run FEM or CFD, assemble the data matrix.
Train
Prepare data, choose a model type, train the model.
Refine
Close data gaps, quantify influencing factors.
Predict
Assess new variants, and inversely: which input reaches the target?
dAIve Predict
From data to prediction. Without code.
Our software for engineers: from data to a validated model, without code. In the browser or on premise.
- Models
- Random Forest, XGBoost, FNN, RNN
- Data
- Simulation, test, process (tabular format)
- Runs
- Browser or desktop (local)
- In use
- with pilot customers since 2023
Who we work for
Our clients are based in Germany, Austria and Switzerland, and we are glad to work internationally. What they share: products that are simulated, measured and tested.
From practice
Accelerating R&D Processes with AI-Driven CAE
With PDTec: how centrally managed simulation data, task-specific AI models and SPDM integration speed up development in the automotive industry.
NAFEMS ↗ · membersConference paper · NAFEMS World Congress 2025Optimized EV Charging: AI-Driven Model Reduction
With AUDI: an AI-reduced model sizes charging cables with phase-change material for fast charging, without simulating every variant.
NAFEMS ↗ · membersNAFEMS SPDM & AI/ML Conference 2026 · Garching near MunichFrom Simulation Data Management to Engineering Intelligence: AI-Integrated Requirement Robustness Assessment (with CAIQ)
AI-Driven Engineering Intelligence for Physics-Based Decision Support in CAE Workflows (with DLR and PTB)
→ InsightsOne group, two locations
One team at two locations, with one shared set of services.
Consulting, training and the dAIve Predict software
Engineering and R&D projects: simulation, hybrid models, product development
Let's talk about your development.
In a first call we clarify where you stand and which step makes sense. No obligation, to the point.