Inference Technologies
Corporate brand
UnclaimedCzech AI/ML software company delivering predictive manufacturing systems and data analytics for semiconductor fabs.
Last updated September 28, 2026
AI crawler reads
No AI crawler has read this profile yet.
Overview
Inference Technologies is a Czech software company founded in 2015 and headquartered in the Czech Republic, with a representative presence in Taiwan. It operates in the semiconductor manufacturing data landscape, building predictive systems and dynamic data analysis tools that are designed to meet high security standards. Its core commercial offering is the DeepFab technology, a suite of AI/ML-based predictive manufacturing systems that work together to improve wafer cost, reliability, and yield. The company positions itself around the idea that automated data collection has become the semiconductor industry standard, and that the next step is an AI standard. Its DeepFab product family includes inProbe for smart product routing, inSure for smart reliability control, inLab for advanced analytics, and inMap for wafer bin map pattern recognition. These modules are intended to deliver reduced wafer cost and required electrical test capacity, reliability improvement and process feedback, root-cause analysis of yield excursions, and pattern-recognition-based monitoring and reporting. Inference Technologies emphasizes data control and security. It describes a pipeline that goes from raw files to model inference using a self-hosted data pipeline, with machine learning models designed to optimize cost and reliability objectives, and deployment that keeps data under the customer's control with model inference behind the customer's firewall. This self-hosted, firewall-first approach is a key part of its value proposition for semiconductor manufacturers that handle sensitive process and yield data. Beyond DeepFab, the company offers SemiGuide, an interactive analytical tool for semiconductor education and research. SemiGuide allows users to explore materials such as Si versus SiC and GaN and to manipulate physical properties to investigate performance and efficiency based on solid-state physics, targeting students, educators, and engineers in semiconductor development and analysis. The company markets its solutions through a consultative, pilot-oriented sales approach. Its pricing page invites prospective customers to discuss what a fab can gain from predictive systems, to run a pilot study, or to adopt a complete DeepFab solution, with pricing depending on the type and scope of services and support. Inference Technologies also participates in industry events; its news section highlights participation in Silicon Saxony Days 2026. The company recruits technical staff such as data engineers and machine learning engineers in Prague, Czechia, and describes a culture of creative freedom, personal development, and time flexibility.
History
Inference Technologies was established in 2015 and is headquartered in the Czech Republic, with a representative in Taiwan. The company was founded to address the semiconductor manufacturing data landscape, focusing on predictive systems and dynamic data analysis solutions that adhere to high security standards. Its core technology, DeepFab, was developed as a suite of predictive manufacturing systems that work together. The product family includes inProbe for smart product routing, inSure for smart reliability control, inLab for advanced analytics, and inMap for wafer bin map pattern recognition. The company frames its mission around the transition from automated data collection, which it describes as already the semiconductor industry standard, to an AI standard. Inference Technologies emphasizes a self-hosted data pipeline that goes from raw files to model inference, with machine learning models designed to optimize cost and reliability objectives. Deployment is designed to keep data under the customer's control, with model inference behind the customer's firewall. This approach targets semiconductor manufacturers that need to protect sensitive process and yield data. The company also offers SemiGuide, an interactive analytical tool for semiconductor education and research. SemiGuide lets users explore materials such as Si versus SiC and GaN and manipulate physical properties to investigate performance and efficiency based on solid-state physics, serving students, educators, and engineers. Inference Technologies markets its solutions through a consultative approach, inviting prospective customers to discuss potential gains from predictive systems, run a pilot study, or adopt a complete DeepFab solution, with pricing depending on the type and scope of services and support. The company recruits technical talent in Prague, Czechia, including data engineers and machine learning engineers, and promotes a culture of creative freedom, personal development, and time flexibility. It also participates in industry events, such as Silicon Saxony Days 2026.
- 2026
Company announces participation in Silicon Saxony Days 2026.
- 2015
Inference Technologies is established and headquartered in the Czech Republic.
Company develops DeepFab, a suite of AI/ML predictive manufacturing systems for semiconductor fabs.
Inference Technologies offers SemiGuide, an interactive analytical tool for semiconductor education and research.
Products and positioning
AI/ML predictive manufacturing systems for semiconductor fabs, with self-hosted data pipelines and model inference behind the customer's firewall for high security standards.
DeepFab
A suite of predictive manufacturing systems for semiconductor fabs, comprising inProbe, inSure, inLab, and inMap.
inProbe
Smart product routing module within DeepFab.
inSure
Smart reliability control module within DeepFab.
inLab
Advanced analytics module within DeepFab.
inMap
Wafer bin map pattern recognition module within DeepFab.
SemiGuide
Interactive analytical tool for semiconductor education and research, allowing exploration of Si vs SiC and GaN and manipulation of physical properties.
Sources3 sources
Cite this profile: Cite the canonical profile. /brand-wiki/inference-technologies · Editorial policy · How profiles are compiled