Are you looking for a cutting-edge DeepTech startup specialising in advanced AI systems, agent-based architectures, and knowledge-driven intelligence?
Our mission is to revolutionise the way data is analysed, connected, and reasoned about in the field of cybersecurity by combining state-of-the-art deep learning, AI agents, and knowledge graph technologies. As a pioneer in this space, we are looking for highly skilled and passionate Deep Learning / AI Scientists to join our team and contribute to our ongoing research and development efforts.
As a Deep Learning Scientist with a focus on AI agents and knowledge-based systems, you will play a key role in the development and implementation of intelligent architectures that combine large language models, autonomous agents, and structured knowledge representations.
You will work closely with a team of talented researchers, engineers, and cybersecurity domain experts to design, train, and optimise AI systems capable of reasoning over large volumes of unstructured and structured data. This role offers an exciting opportunity to contribute to cutting-edge research, impact real-world applications, and shape the future of AI-driven cybersecurity intelligence.
Research and Development
Conduct state-of-the-art research in deep learning, agent-based systems, and knowledge-enhanced AI. Explore novel architectures such as Transformer-based models (e.g. BERT, GPT), retrieval-augmented generation (RAG), multi-agent systems, and neuro-symbolic approaches to improve reasoning, planning, and decision-making capabilities.
Agent and System Design
Design and implement AI agents that collaborate, plan, and reason over complex problem spaces. Develop architectures that integrate LLMs with tools, memory, feedback loops, and structured knowledge sources.
Knowledge Graphs and Representation Learning
Develop and apply techniques for building, maintaining, and leveraging knowledge graphs, ontologies, and semantic representations. Combine symbolic knowledge with learned representations to enhance explainability, consistency, and reasoning performance.
Data Preprocessing and Feature Engineering
Develop advanced techniques for preprocessing and representing both unstructured text and structured data, including tokenisation, embeddings, entity linking, relation extraction, and graph-based representations.
Model Training and Evaluation
Train and fine-tune deep learning models using techniques such as transfer learning, self-supervised learning, in-context learning, and reinforcement learning for agents. Evaluate system-level performance using appropriate metrics and propose improvements for robustness and scalability.
Collaborative Research
Work closely with cross-functional teams including data scientists, engineers, and cybersecurity specialists to understand real-world requirements, contribute to project planning, and translate research into production-ready systems.
Documentation and Reporting
Document research findings, system architectures, and experimental results clearly and concisely. Prepare technical documentation, internal reports, and presentations for both technical and non-technical audiences.
Cyber Insight is revolutionizing the IT security and regulatory markets with an accessible SaaS solution that enables law firms, IT service providers, management consultants, insurance companies and their clients to identify, manage and mitigate risks.
Cyber attacks cause a total annual damage of 223 billion euros in the German economy, with 88% of companies being affected by malicious attacks. According to our evaluations, a data protection breach costs small and medium-sized companies an average of around €67,000.
The Cyber Insight Platform helps our customers to increase the compliance step by step. Compliance includes both security and data protection standards. The assessment of risks and the derivation of prioritized recommendations for action are based on a multifactorial risk model that takes into account individual vulnerabilities and exogenous factors such as current trends or changes in the legal situation and integrates them automatically into the assessment scheme with the help of AI. The customer thus receives automated security assessments and suggestions for concrete measures, giving them the opportunity to reduce their risk cost-effectively and mitigate damages and penalties before they occur.
The goal is to make cyber risks and regulatory compliance measurable and to enable anyone to adequately address these issues who would previously have been exposed to very high risk due to insufficient capitalization or lack of know-how.
