Quy mô công ty : 10-24
Xem thêmTOG recruitment
For one of our clients, TOG recruitment is looking for a Senior Developer to work in Ho Chi Minh City. As such, you always will be working with the latest proven technologies in a state of the art office with an open environment. Our client is an IT Staffing company specializing in software and application development. They provide technology services that digitally transform their partners and generate business value. They are a Microsoft Gold Partner and are ISO 27001:2013 certified. They offer a comprehensive range of services, including staffing and technology consulting, software engineering, and ongoing support. They fully embrace the Agile software engineering principles and use SCRUM to manage projects. Their multidisciplinary expertise enables them to deliver innovative solutions that help their partners achieve their goals.
Key Responsibilities:
• Lead the design and implementation of AI-powered features using LLM APIs (OpenAI, Anthropic Claude, Google Gemini) including tool use, function calling, and multi-step chaining.
• Architect and build RAG (Retrieval-Augmented Generation) pipelines integrated with vector databases such as Pinecone, Weaviate, or pgvector.
• Define and enforce prompt engineering standards — designing reusable prompt templates, evaluation frameworks, and guardrails for production AI workflows.
• Champion the use of AI coding tools (GitHub Copilot, Claude Code) across the engineering team, driving productivity and code quality improvements.
• Build and maintain internal AI workflow automation using Azure AI Foundry, including model deployment, evaluation, and prompt flow management.
• Leverage Microsoft AI stack — Azure OpenAI Service, Azure AI Search, and Semantic Kernel for orchestration.
• Evaluate and recommend emerging AI developer tools, APIs, and platforms to keep the team ahead of the curve.
• Conduct code reviews with a focus on AI integration patterns, security, and cost efficiency (token usage, API rate limits).
• Mentor and upskill engineers on responsible and effective AI-assisted development practices.
• Collaborate with product managers to translate business requirements into AIenabled technical solutions
• 5–8 years of software engineering experience, with at least 2+ years focused on AI tool integration and LLM-powered application development. (Must have)
• Hands-on expertise with Azure OpenAI Service — GPT-4o, embeddings, streaming, function calling, and system prompt design. (Must have)
• English level required: Advanced or Fluent, able to communicate effectively with international stakeholders (Must have)
• Proven experience with Azure AI Foundry — model catalog, prompt flows, evaluation pipelines, and AI project management. (Must have)
• Deep proficiency in prompt engineering: few-shot prompting, chain-ofthought, structured output, and evaluation techniques. (Must have)
• Hands-on experience with Microsoft Semantic Kernel for AI orchestration and agent workflows. (Must have)
• Practical experience with AI coding assistants, especially GitHub Copilot and Claude Code, and driving team-wide adoption. Proficiency in TypeScript/JavaScript and .NET. (Must have)
• Solid experience with Azure cloud services (App Service, Functions, Key Vault) and modern CI/CD on Azure DevOps or GitHub Action. (Must have)
• Experience with other LLM APIs: Anthropic Claude, Google Gemini, or Mistral. (Nice-to-have)
• Familiarity with open-source AI frameworks such as LangChain, LlamaIndex, or the Vercel AI SDK. (Nice-to-have)
• Experience with AI observability and evaluation tools (LangSmith or Azure AI Foundry Evaluation (Nice-to-have)
• Exposure to multi-agent frameworks (AutoGen, Semantic Kernel Agents, or LangGraph). (Nice-to-have)
• Background in developer tooling, platform engineering, or DevOps on Azure. (Nice-to-have)
English
Nói: Advanced - Đọc: Advanced - Viết: Advanced
By utilizing cutting edge technology and our deep knowledge of IT we are able to deliver the best suitable matches within agreed time. Not only limited to the higher functional levels but for all.
