Senior AI Developer (Construction Technology)

We are seeking an AI Developer to design, build and deploy intelligent systems that enhance construction operations, safety and productivity across multilingual environments.

Responsibilities:

- Develop and deploy AI-powered autonomous agents and multi-agent orchestration systems for construction project management, using large language models and agentic loop architectures

- Build and maintain end-to-end speech technology pipelines (automatic speech recognition and text-to-speech synthesis) enabling real-time multilingual voice interfaces for field workers and customers, including model deployment on GPU infrastructure, latency optimisation and failover management

- Design and implement retrieval-augmented generation (RAG) systems and automated data management frameworks for construction document repositories (permits, contracts, safety manuals, etc.) with semantic search

- Develop custom MCP (Model Context Protocol) server integrations connecting AI agents to internal construction management databases, IoT sensor feeds and ERP systems

- Manage GPU compute infrastructure for AI model serving, including CUDA environment management, model quantisation, vRAM optimisation and container orchestration

- Research and implement automated CI/CD pipelines for AI model deployment, A/B testing and performance monitoring in production environments

- Create real-time data visualisation dashboards and reporting tools for construction project KPIs powered by AI-driven analytics

- Evaluate and benchmark emerging AI models, maintaining internal documentation on model capabilities, cost-performance trade-offs and deployment readiness assessments

- Provide external AI development consulting services to third-party clients, including requirements analysis, system architecture design, proof-of-concept development and production deployment of custom AI solutions

- Conduct technical feasibility assessments and prepare proposals for client AI projects, covering scope definition, technology stack recommendation, timeline estimation and cost-benefit analysis

- Deliver end-to-end AI system integration for external clients: from system design through development, testing, deployment and post-launch support, with a focus on LLM-powered automation, conversational AI and speech-enabled interfaces

- Represent the company in client-facing technical discussions, pre-sales demonstrations and industry events related to AI and construction technology

- Develop reusable AI solution templates and reference architectures that can be customised for different verticals (construction, property management, facilities maintenance)

Requirements:

- PhD in Computer Science, Electrical Engineering, or closely related field

- Minimum 3 years of hands-on experience building and deploying production AI systems (research prototypes alone do not qualify)

- Demonstrated expertise in BOTH automatic speech recognition (ASR) AND text-to-speech (TTS) system deployment, including model training, fine-tuning, and end-to-end speech pipeline engineering

- Strong proficiency in programming languages such as Python, frameworks such as PyTorch, LangChain, Linux and macOS development environments

- Experience with LLM agent frameworks, prompt and context engineering, harness for agentic AI systems, and real-time system integration

- Proven track record of deploying AI services on bare-metal GPU servers (not just cloud-managed services), including NVIDIA driver management, CUDA toolkit, conda/venv environment isolation

- Working proficiency in both English and Mandarin Chinese (written and spoken) for bilingual system development and communication

- Self-motivated, enthusiastic in research on AI-related fields

- Excellent client-facing communication skills with ability to translate complex AI concepts into actionable business recommendations for non-technical audience

Nice to have:

- Published research in speech processing, NLP or AI-related conferences (Interspeech, ICASSP, ACL, EMNLP, etc.)

- Experience with model post-training, knowledge distillation or model quantization techniques

- Prior exposure to construction, civil engineering or built-environment technology