With KnowFab, we are starting in Leipzig. We are building industrial AI for engineering, quality, and production with a clear commitment to transparency instead of black-box decisions.
What drives us
Manufacturing creates large amounts of process and quality data every day. Even so, many decisions still depend on experience, manual analysis, and fragmented systems. That is the gap we want to close.
Many conventional AI solutions detect patterns but provide too little technical reasoning. That is not enough for engineering, QA, and production. When a model recommends something, the path behind that recommendation needs to stay understandable.
AI in production must not be a black box. Every decision needs a path that an engineer can follow.
Tom Suesskind, Co-Founder & CEO
Our approach
KnowFab combines data-driven models with explicit process knowledge. Neural networks and structured knowledge models do not operate side by side, but together. The result is output that is statistically strong and technically interpretable.
Neural network, knowledge graph, technical framing
Our approach combines pattern recognition with explicit process knowledge. This turns production data into an assessment that is technically robust and explainable.
reads unstructured process data and creates initial activation values.
organizes these values along permitted relationships and process logic.
produces a robust assessment with technical context instead of a black-box score.
We combine pattern recognition with structured process knowledge. That is how explainable industrial AI for joining processes becomes practical.
Hong Li, Co-Founder & CTO
What we are launching with
KnowFab Design
Digital support for planning, evaluation, and rule checks in production and joining processes.
KnowFab JoinTech
Analysis of live manufacturing processes, early deviation detection, and support for root-cause analysis.
Pilot projects
Additional development projects help us build references and refine the products close to real industrial requirements.
What makes us different
Focus on joining processes
We concentrate on exactly those process areas where quality risk and operational leverage are especially high.
Knowledge graph as structure core
Process knowledge, material relations, and quality logic are modeled explicitly and remain understandable.
No black-box decisions
Engineering, QA, and production teams get reasoning instead of abstract scores and can act on results with confidence.
Why explainable AI matters in manufacturing
In industrial practice, a prediction alone is rarely enough. When teams evaluate quality deviations, material effects, or process windows, they need to understand which signals influenced the result and which technical relationships sit behind it. That is where trust, release readiness, and real adoption begin.
For engineering, QA, and production teams, this means fewer discussions around opaque scores, faster root-cause analysis when deviations occur, and a better basis for standardized decisions. For us, explainable AI is therefore not an extra feature. It is a requirement for making digital assistance operationally useful on the shop floor.
Who we are
Tom Suesskind
Leads company management, company building, and customer development at KnowFab. Connects market requirements with operational execution, partnerships, and clear day-to-day structures.
Hong Li
Industrial AI, knowledge graphs, and robust architectures for edge and cloud systems.
Jan Paul Buchwald
Software architecture, B2B markets, and SaaS/cloud strategy for industrial products.
Why Leipzig
Our location is a deliberate choice: Leipzig is close to industrial networks in automotive and manufacturing while also offering access to research, talent, and a strong technology environment.
What we are focusing on in year one
- Pilot projects with a clearly defined technical benefit and a solid data basis
- Product maturity for KnowFab Design and KnowFab JoinTech along real manufacturing requirements
- Reference projects in automotive, supplier networks, and adjacent manufacturing environments
- Reusable methodology for explainable assessments instead of isolated one-off solutions
Our goal is a controlled start with clear priorities. We do not want to touch many use cases at once. We want to solve a few highly relevant problems in a way that creates reliable references, robust product building blocks, and clean integrations. That focus is part of how we define quality.
What comes next
With KnowFab, the next step is building partnerships and first references. We are looking for companies in industry and automotive that want to bring explainable AI into real operations with us.