Senior AI Engineer - Model Architecture & Performance (m/f/d)

About Mercury.ai

We are a growing B2B SaaS company from Bielefeld, offering a leading platform for Conversational AI. While conventional language models are often uncontrollable, Mercury.ai relies on Mercury Intelligence: an orchestrated AI structure that uses specialized models for individual tasks to deliver precise and stable results at an enterprise level. Our customers include prestigious companies such as Volkswagen Bank, Nestlé, and Ravensburger.

Senior AI Engineer - Model Architecture & Performance (m/f/d)

In the enterprise environment, two factors are crucial: precision and speed. You will be responsible for further developing our model architecture so that it exhibits minimal latency even for complex tasks (e.g., through the use of diffusion models). You will bridge the gap between theoretical cutting-edge research and high-performance production.

Your Tasks

• You will further develop and optimize our proprietary model structures to continuously increase the response quality and control of Mercury Intelligence.

• You will evaluate and implement modern approaches (such as diffusion models) for specific generative challenges within our Conversational platform.

• You will be responsible for performance engineering, ensuring that our models are optimized for real-time deployment in high-frequency customer channels (inference speed, quantization, caching strategies).

• You will work on the independence of our systems and ensure that our AI infrastructure delivers peak performance without relying on external, privacy-critical third-party APIs.

• You will bring current insights from research directly into product development and support the transfer of prototypes into stable enterprise operations.

Your Profile

• You hold a degree or a PhD in Computer Science, Mathematics, or a field focusing on Artificial Intelligence / Machine Learning.

• You bring several years of experience in the development and deployment of AI models in production environments.

• You have in-depth knowledge of model architectures (Transformers, Diffusion) as well as Java.

• You have a passion for system efficiency and understand how to harden models for low-latency scenarios.

• You share our commitment to data protection (GDPR-compliant) and the vision of an independent European AI landscape.

What We Offer

• Work on a platform that does not depend on OpenAI & Co., but operates technologically autarkic.

• An environment rooted in Bielefeld University that combines scientific curiosity with the momentum of a successful B2B SaaS company.

• A stylish office in downtown Bielefeld with a rooftop terrace, where we solve technical problems together at the whiteboard.

• Your developments will drive communication for millions of users at global market leaders and DAX corporations.

Your Way to Us

👉 Do you want to shape the next-generation architecture of Conversational AI? Send us your application documents including relevant publications or projects to hi@mercury.ai