Software Engineer, GeminiApp Info Seek Quality

Snapshot We are the Information Verticals team for the Gemini App, bringing search, maps, shopping and other information-related features to both the web and mobile Gemini applications. We are a fast-paced organization bringing features from ideation to launch in a matter of months. We are a non-bureaucratic, innovative, bottom-up organization. We are seeking an experienced and self-motivated Software Engineer with particular expertise in quality and AI applications to ensure the Gemini applications deliver world-class, reliable, and personalized experiences. About Us Within the Information Verticals team, the Shopping team is at the forefront of Google's shift towards agentic experiences. We are redefining how users discover, research, and purchase products by combining the reasoning capabilities of Gemini with the comprehensive data of the Shopping Graph. To constantly innovate, we work closely with various Google DeepMind (GDM) research units. Following a breakout year with two major launches in Q4 ( launch announcement ) that were well received in the press, we have established a new baseline for highly visual agentic shopping. We have one of the highest favorability retention metrics on GeminiApp have and were featured prominently in the recent "Personal Intelligence" launch ( launch , Sundar tweet ). This role offers the unique opportunity to make impact on a high-growth vertical that is already seeing massive traction,). The Role Join Google DeepMind's mission to advance science and benefit humanity by solving intelligence. We are seeking an experienced and self-motivated Software Engineer with a passion for quality to ensure the Gemini applications deliver world-class, reliable, and personalized experiences. In this role, you will be building new features in the Gemini product, focused on shopping and other information-related verticals like search, maps news. You’ll prototype new LLM applications, design and implement evaluation metrics, collection evaluation data, analyze model responses, and conduct prompt engineering in order to ship high-quality features to our users. This is a rare opportunity to shape a new era of technology, accelerating development with the latest AI models and collaborating with AI to build high-fidelity user experiences. You will see your work in the hands of users frequently and quickly. As a key contributor to one of Google's top efforts for 2025, you will help shape the future of AI. What You'll Do Implement sophisticated LLM-powered systems, including context engineering , agentic workflows with tool use, and robust evaluation frameworks. Dive deep into model behavior, creating high-quality, diverse, and challenging evaluation datasets to identify weaknesses and guide improvements. Make improvements to the serving and evaluation infrastructure to power new features. Develop and scale evaluation solutions (auto-raters, human evals) to continuously monitor and improve the performance of LLM-powered features. Drive collaboration with cross-functional partners (Product, UX, Research) to translate user needs into intuitive and valuable features within the Gemini ecosystem. Collaborate with GenAI research teams to integrate product-driven data and evaluations to the base Gemini models. Analyze user data and model outputs to enhance the quality of personalized experiences, ensuring Gemini is helpful and relevant for every user. Champion a culture of quality, defining metrics, setting standards, and driving best practices for LLM evaluation and responsible AI development. About You We seek out individuals who have the skill and passion to design and build high-quality components, while simultaneously being willing to move fast and pivot when necessary. We are looking for strong and adaptable engineers who are obsessed with quality and have a deep understanding of how to measure and improve AI systems. You are a flexible problem-solver, comfortable inventing novel solutions ... (truncated, view full listing at source)