Gradion AI

We are an IT consulting firm specializing in machine learning and agentic AI projects. Combining extensive expertise in ML and AI [1][2][3], with decades of experience in software development and systems engineering [1], we create custom solutions that are intelligent, meet customer requirements and run reliably in production.

In-house, we are currently developing a group of AI agents collaborating with us to run our business. Our focus is on agents that collaborate with entire teams, rather than just individual users. To be most useful for team members, we provide agents with the full context of team dynamics. We also research collaborative and autonomous skill learning in code action agents.

hybrid-groups[↗] integrates human team collaboration with AI agents. Unlike typical AI applications built for single-user interaction, hybrid-groups enables AI agents to have conversations with multiple users simultaneously, recognizing each user’s identity and respecting their unique preferences and permissions. Agents can also proactively support team discussions by contributing relevant information or initiating helpful actions.

freeact[↗], a lightweight AI agent library using Python as the common language to define executable actions and tool interfaces. This is in contrast to traditional approaches where actions and tools are described with JSON. A unified code-based approach enables freeact agents to reuse actions from earlier steps as tools or skills in later steps. This design allows agents to build on their previous work and compose more complex actions from simpler ones.

ipybox[↗], a lightweight and secure Python code execution sandbox based on IPython and Docker. You can run it locally on your computer or remotely in an environment of your choice. While optimized for AI agents that interact with their environment through code execution, such as freeact agents, ipybox also serves as a general-purpose sandbox for secure code execution.

MerlinOne[↗]. Conception and development of a multimodal agentic AI search engine with semantic understanding of images, videos, documents, and audio, enhanced through domain-specific model fine-tuning for media organizations. In addition to many other production deployments, it powers the Associated Press Newsroom, and contributed to MerlinOne's successful acquisition by Canto in 2023.

Canto[↗]. Development of an AI-powered visual and hybrid search platform built as a cloud-native solution that scales seamlessly to thousands of enterprise customers worldwide. Training of custom query processing models with synthetic data to deliver highly accurate search results across diverse content types.

cyan Security Group[↗]. Design and implementation of an end-to-end MLOps platform for the entire model lifecycle, automating data pipelines, model training, and deployment workflows. Fine-tuning of custom models for content filtering and child protection to enhance digital safety at major telecom providers.

Red Bull Media House[↗]. Global distribution of the in-house DAM system to support low-latency local access to content and write-availability under network partitions. Release of the underlying algorithms and protocols as open source project. Content playout platform for Red Bull TV's global streaming service, delivering content to millions of users worldwide with consistent performance.

Martin Krasser[↗]. ML and AI engineer with strong background in deep learning, agentic systems, distributed systems, and system integration. Specialized in developing reliable AI systems and operating them at scale in production environments. Extensive industry experience in both technical and leadership roles. Active open source contributor.

Christoph Stumpf[↗]. ML and AI engineer with strong background in MLOps, agentic systems, distributed systems, and cloud-native architectures. Proficient in engineering reliable AI applications and supporting their entire lifecycle. Comprehensive professional expertise in building MLOps platforms and mentoring teams. Active open source contributor.