5 of 5 Permanent Benchmarking Jobs in Central London

frontier relevant to agentic systems e.g., LLMs, multimodal models, structured reasoning, data generation, or video-text modelling. Build robust training, evaluation, and benchmarking setups for new model variants and agent behaviours. Prototype new agent capabilities and work closely with engineering teams to turn them into production-quality features. Analyse failure ...

At Forvis Mazars, we're always looking ahead, for our people, our clients and the wider world. Together, we grow , belong and impact . You'll be supported to learn, explore and develop from day ...

maintain clear architecture documentation and technical roadmaps. - Identify improvement opportunities and drive them through design, build, and deployment. 2. Establish Robust Metrics, Monitoring & Benchmarking - Define and track KPIs across performance, reliability, feasibility, and commercial impact. - Build back-testing, scenario analysis, and version benchmarking frameworks. - Create monitoring tools that support rapid ...

Google Cloud Vision AI, and Azure AI Services, while cultivating relationships with emerging AI and edge startups. Your partnerships will drive integration quality, benchmarking, SDK compatibility, and technical validation that compound our long-term advantage. Operating as both strategist and operator, you’ll collaborate with Product and Engineering on integrations … benchmarking, work with Marketing on co-marketing and joint content, and partner with Sales/Customer Success on enablement that increases customer value. You’ll represent Ultralytics at flagship events such as YOLO Vision, GTC, and CVPR — building thought leadership through talks, panels, and joint announcements. Success looks like measurable ...

research. You’ll play a key role in developing and refining new ML architectures, designing input/output encoding methods, and conducting rigorous benchmarking to evaluate and improve performance. This role is ideal for an engineer with a strong foundation in ML who enjoys both conceptual design and hands … software ML models tailored to next-generation computing platforms. Design and evaluate data encoding, preprocessing and postprocessing methods. Build tools and pipelines for performance benchmarking and model evaluation. Research and adapt state-of-the-art ML and computer vision techniques to new hardware paradigms. Collaborate cross-functionally with hardware, software ...