Monika Grajzer

I help companies design and evaluate complex AI systems, combining applied research expertise with practical system architecture designs.

With a Ph.D. and 15+ years of experience in applied R&D and artificial intelligence, I specialize in building custom AI and machine learning solutions for technically challenging problems where off-the-shelf tools are not sufficient.

My work often focuses on the early stages of AI initiatives — helping organizations assess feasibility, design system architecture, and avoid costly technical mistakes before large investments are made.

I support clients in the areas such as:

- AI feasibility and technical risk assessment

- architecture design for AI and machine learning systems

- technical due diligence for AI-driven startups/investors

- evaluation and improvement of existing AI solutions

- designing custom ML pipelines and data processing systems

- intelligent monitoring systems, including network management and traffic analysis

Over the years I have developed several AI / ML / Deep Learning solutions such as:

- conversational AI systems and virtual avatars (Voice Agents)

- voice biometrics (speaker recognition) and keyword spotting

- solutions for mobile/embedded devices

- speech feature extraction and speech processing pipelines

- large-scale data processing pipelines for automated content generation and normalization

- automated network management and network traffic analysis

I have built solutions where ML models were deployed both in the cloud and directly on devices (including Tensorflow Lite), integrating models and APIs from multiple vendors (OpenAI, Anthropic (Claude), DeepSeek, etc.) including real-time conversational systems using OpenAI Realtime API, Agents SDK and Google Multimodal Live API.

I also have hands-on programming experience in Python, Kotlin (Kotlin multiplatform), Java, C/C++, and Node.js which allows me to prototype solutions, evaluate implementations, and work effectively with development teams.