Generalist software engineers. Much of Google's hiring is for one role, the software engineer, who is expected to write code in whichever of Java, C/C++, Python, JavaScript, TypeScript or Go a team uses, and to move between teams and products. These postings name no product and no single stack. They require one or more of those languages, testing, and experience in one of a list of broad areas, information retrieval, the science of finding the right result in a vast store of documents, and machine learning above all. Some lean to building applications, with HTML and CSS for the front end, some to infrastructure, and some to backend services with SQL and Spring Security. Together they are much the largest part of Google's hiring, and their share grew from October to December 2025 to July to September 2026. This is also the role that takes juniors most readily.

C/C++ systems engineers. Google runs one of the largest private networks and fleets of data centres in the world, and builds its own compilers and system software for them. C/C++ systems engineers on Linux build that layer. The postings require C/C++ with compilers, the Linux kernel and device drivers, and some require networking in depth, Ethernet, TCP/IP, routing at Layer 2 and Layer 3, quality of service, NAT, firewalls, BGP and DPDK, a toolkit for moving network packets at very high speed. Many also require security knowledge. These engineers are more than twice their usual weight, mostly senior, and their jobs stay open far more often than usual.

ML engineers. Google hires machine-learning engineers at many times their weight across product companies. Most are model-centric ML and data-science engineers, who work on language and on speech and audio, and a smaller group train and adapt models. The postings require Python, often with C/C++, data processing and Google Cloud, and some require JAX and TensorFlow, Google's own frameworks for building models. Many name speech and audio, reinforcement learning, where a model learns by trial and reward, or information retrieval among the fields they want experience in. These roles take juniors about twice as readily as ML roles at product companies. Their share of openings has fallen in every quarter of the record, from one of the largest roles here to a much smaller one.

GenAI engineers. Gemini, Google's own AI model, is going into Search, which increasingly writes its answer rather than listing links, and into Android, Workspace and Google Cloud. GenAI engineers build those features and the agents that act on a user's behalf. The postings require Python, often with Java and JavaScript, with information retrieval, Vertex AI, Google Cloud's service for building with models, and ADK, Google's kit for building agents. Some require Gemini itself, spoken conversational AI, or the newer protocols that let agents use tools and talk to one another, MCP and A2A. GenAI engineers are well above their usual weight, mostly senior, and their share of openings rose sharply in July to September 2026.

Embedded firmware engineers. Google makes Pixel phones, Chromebooks and its own chips for its data centres, and Android runs on billions of devices made by others, like Xiaomi and Motorola. Embedded engineers write the software that runs directly on that hardware. The postings require C/C++ and assembler with systems on a chip, ARM and x86, the Linux kernel, device drivers and board support packages, real-time operating systems and bootloaders, and a smaller group work on the Android platform itself. Embedded engineers are at about their usual weight, but their share of openings grew in every quarter of the record.

Software Engineer, Security. Software engineers who specialise in security: they build login systems, cloud and network protections, and the checks that keep code safe. Google sells security products through Google Cloud and has to keep Search, YouTube and Android safe. Most security software engineers here are application security engineers, with some network security engineers and cloud security and DevSecOps engineers. Some postings require Linux, and some require Prisma Cloud, a tool for checking that cloud systems are set up safely. They are a little above their usual weight, and their jobs stay open unusually often.

Also present. Mobile engineers, at about their usual weight, build Android apps in Kotlin and Java. Cloud and DevOps engineers, well below their usual weight, require Go, Python and Java with Google Cloud. Java engineers are a small fraction of their usual weight, and many Java postings here also require Go, Python and TypeScript. Data engineers, mostly working with databases, and frontend engineers in Angular and TypeScript are well below their usual weight. Full-stack engineers appear at a little below theirs. A small group of data analysts in Hyderabad build business intelligence in Python, many times their usual weight. A few Go backend engineers, chip designers working with EDA tools, fintech software engineers working on mobile payments in Kotlin and Swift, introduced as Software Engineer, Fintech, and test engineers, a small fraction of their usual weight, complete the picture.

Rare here. Python backend engineers, .NET engineers, and SAP, Salesforce and ServiceNow engineers barely appear.