LLM Platform Engineer – Gemini App Skills

DeepMind Technologies Limited; Mountain View (CA); On-site; USD 174,000 - 252,000; Full time

Job summary

DeepMind is seeking a Software Engineer for the Gemini App Skills Platform in Mountain View to help transform Gemini into an action-oriented, universal agent. You will architect systems for discovering, retrieving, and executing modular capabilities across Google, DeepMind, and third-party ecosystems, bridging foundation models with real-world execution.

Your work will involve production serving infrastructure (C++) and Python evaluation pipelines, focusing on low-latency, scalable design and

Qualifications

• Experience developing software with large language models, agentic workflows, tool use, or retrieval-augmented generation (RAG).
• Experience designing evaluation datasets, automated rating systems, or conducting error analysis and prompt optimization for generative models.

Responsibilities

• Design and build intelligent skill discovery, hybrid routing architectures, and low-latency C++ serving components across Gemini on web, mobile, and Chrome.
• Pioneer dynamic system instructions, prompt representations, and semantic retrieval schemas enabling scalable skill catalogs.
• Design automated evaluation frameworks to measure triggering precision, multi-turn tool use, and parameter validity.
• Perform loss analysis on production telemetry and guide prompt/model improvements for reliability and architecture quality.
• Collaborate with DeepMind Research, Search, Vertical teams and external engineers to onboard modular capabilities.

Skills

LLM development

RAG / tool use

Evaluation design

Prompt optimization

Tools

C++

Python

Job description

DeepMind is seeking a Software Engineer for the Gemini App Skills Platform in Mountain View to help transform Gemini into an action-oriented, universal agent. You will architect systems for discovering, retrieving, and executing modular capabilities across Google, DeepMind, and third-party ecosystems, bridging foundation models with real-world execution.

Your work will involve production serving infrastructure (C++) and Python evaluation pipelines, focusing on low-latency, scalable design and