(Senior) AI Engineer

A senior AI Engineer will join a rapidly expanding AI‑focused product firm that delivers a forward‑looking digital assistant used by millions worldwide. The position centers on shaping the core backend that powers the assistant’s capabilities across mobile and desktop platforms. The day‑to‑day work involves taking full responsibility for the AI serving layer, including inference endpoints, agent runtime, tool connectors and public APIs. You will design and run fast, high‑capacity services that translate model outputs into reliable production features. Building and refining agent pipelines that manage planning, multi‑step tool usage, error detection and recovery is a key part of the role. You will also evolve the machine‑learning platform by creating model‑serving infrastructure, deployment pipelines, benchmarking tools and experiment frameworks for researchers and engineers. Implementing thorough observability, monitoring, tracing and alerting, as well as leading incident response, will be essential. Continuous optimisation of response time, throughput, cost and stability across inference, caching, batching, streaming and GPU usage is expected. Clear service boundaries between models, backend components and product layers must be defined to enable rapid iteration, while working closely with model owners, product managers and frontend teams to ensure AI features are performant and aligned with user expectations. The ideal candidate holds a master’s or bachelor’s degree in computer science, engineering or a comparable discipline, or possesses equivalent hands‑on experience. At least five years of experience building and operating production‑grade backend systems, distributed services or AI‑focused infrastructure at scale are required. Deep knowledge of Python and Node.js is mandatory, together with proven work using PyTorch or JAX and LLM serving frameworks such as vLLM, SGLang or TensorRT‑LLM. Proficiency with Kubernetes, Docker, cloud platforms, relational, NoSQL and vector data stores, and modern observability stacks is expected. Strong software‑engineering fundamentals, clean code practices, system design and API architecture are essential. Hands‑on expertise with AI/LLM inference patterns, retrieval‑augmented generation, agent workflows, streaming results and partial outputs is required, as is a track record of delivering low‑delay, high‑capacity services. The role demands a data‑driven mindset, comfort with ambiguous requirements, a strong sense of ownership and excellent communication skills. Full professional fluency in English is a must. Experience with frontend technologies such as Next.js is considered a plus, and the position offers the chance to influence the technical foundation of a next‑generation AI assistant at massive scale. What the role asks for: - Master's or Bachelor's in computer science, engineering or related field - 5+ years building/operating production backend or AI infrastructure - Proficient in Python - Proficient in Node.js - Experience with PyTorch or JAX - Experience with LLM serving (vLLM, SGLang, TensorRT‑LLM) - Knowledge of Kubernetes and Docker - Familiarity with cloud platforms - Experience with SQL, NoSQL and vector databases - Ability to implement observability, monitoring and alerting - Proven design of low‑delay, high‑capacity services - Strong software engineering fundamentals - Hands‑on AI/LLM inference, RAG and agent workflow experience - Full professional proficiency in English - Excellent communication and cross‑team collaboration - Ownership mindset for ambiguous problems - Frontend experience with Next.js (nice‑to‑have)