AI Engineer (LLM/ Chatbot)

Pantheon Lab Limited; Hong Kong; On-site; HKD 480,000 - 750,000; Full time

Job summary

Pantheon Lab Limited seeks an experienced AI Engineer to design, deploy and optimize production-grade language model systems. You will build applications using API-based and self-hosted models, implement RAG pipelines, and create end-to-end LLM workflows.

The ideal candidate combines hands-on LLM API integration with expertise in deploying and optimizing local models for practical use cases. Strong Python, FastAPI, and cloud familiarity are expected.

Qualifications

• 2+ years building production-grade LLM applications.
• Experience integrating commercial LLM APIs.
• Hands-on experience deploying local models in production.
• Familiarity with prompt engineering and evaluation.

Responsibilities

• Design high-throughput, low-latency serving architectures for LLM apps.
• Build and maintain RAG pipelines and end-to-end workflows.
• Integrate commercial LLM APIs into production systems.
• Develop prompt engineering techniques and management systems.
• Deploy and serve local open-source models for specific use cases.
• Optimize local model inference performance with efficient serving frameworks.
• Fine-tune models to improve domain-specific tasks.
• Monitor and troubleshoot production LLM systems; ensure reliability.
• Document architectures, best practices, and decisions.
• Collaborate with engineering teams to integrate LLM capabilities into products.

Skills

Python programming

async/await

transformer models

LLM APIs

local models

RAG pipelines

FastAPI

vector DBs

NoSQL

Docker

LoRA/finetuning

vLLM/TGI

Hugging Face

cloud platforms

Git/GitHub

Tools

LangChain

LlamaIndex

Docker

Kubernetes

MongoDB

Pinecone

Weaviate

Milvus

GitHub

Job description

We are seeking an experienced AI Engineer to manage the design, deployment, and optimization of production‑grade language model systems. This role involves building applications using both commercial LLM APIs and self‑hosted open‑source models, implementing RAG pipelines, and creating end‑to‑end LLM workflows. The ideal candidate combines practical experience integrating LLM APIs with technical expertise in deploying and optimizing local models.

Key Responsibilities

• Design and implement high‑throughput, low‑latency serving architectures for LLM applications
• Build and maintain RAG pipelines and end‑to‑end LLM workflows
• Integrate and optimize commercial LLM APIs (OpenAI, Anthropic, Google, etc.) into production systems
• Develop prompt engineering techniques and prompt management systems
• Deploy and serve local open‑source language models for specific use cases
• Optimize local model inference performance through efficient serving frameworks
• Fine‑tune models to improve performance on domain‑specific tasks (
• Monitor and troubleshoot production LLM systems to ensure reliability
• Research and experiment with emerging models and techniques to improve system capabilities
• Document architectures, best practices, and technical decisions
• Collaborate with engineering teams to integrate LLM capabilities into products
• Communicate technical terms and recommendations to stakeholders
• Build evaluation frameworks to measure model quality, latency, cost, and user satisfaction
• Design intelligent routing and fallback strategies across multiple LLM providers
• Scale LLM services to handle production workloads efficiently
• Implement caching, batching, and request optimization strategies for both APIs and local models

Required Qualifications

Experience:

• Solid 2+ years focused on LLM applications or chatbot development(Candidates with more experience will be considered for a senior role.)
• Proven track record of building production LLM applications
• Experience integrating and optimizing commercial LLM APIs
• Hands‑on experience deploying local models in production environments

Technical Skills:

• Strong Python programming with emphasis on async/await patterns and production‑quality code
• Deep understanding of transformer architectures and LLM fundamentals
• Experience with LLM APIs (OpenAI, Anthropic Claude, Google Gemini, or similar)
• Familiarity with local open‑source models (Qwen, Llama, Mistral, or similar)
• Experience with RAG implementation using LlamaIndex, LangChain or similar frameworks
• Proficiency with FastAPI for building high‑performance APIs
• Experience with vector databases (Pinecone, Weaviate, Chroma, Milvus, or similar)
• Working knowledge of MongoDB or other NoSQL databases
• Experience with Docker containerization and deployment
• Good to have hands‑on fine‑tuning experience (LoRA, QLoRA, full fine‑tuning)
• Familiarity with local model serving frameworks (vLLM, TGI, or similar)
• Familiarity with LLM workflow tracing and observability frameworks (MLflow, Phoenix, Langfuse, or similar)
• Familiarity with Hugging Face ecosystem and transformer libraries
• Experience with cloud platforms (AWS, GCP, or Azure)
• Proficiency with Git/GitHub and version control workflows

Domain Knowledge:

• Understanding of prompt engineering and optimization techniques
• Knowledge of LLM evaluation metrics and benchmarking methodologies
• Experience with cost optimization for LLM applications
• Familiarity with distributed computing and scaling strategies
• Understanding of LLM inference optimization (quantization, batching, caching)

Preferred Qualifications

• Understanding of digital human technologies and multimodal applications
• Knowledge of MLOps practices and CI/CD for ML systems
• Experience with Kubernetes for container orchestration
• Experience with streaming inference and real‑time applications
• Background in function calling and tool use with LLMs
• Familiarity with RLHF (Reinforcement Learning from Human Feedback)
• Experience with model distillation and knowledge compression
• Understanding of distributed training and GPU optimization
• Experience with multi‑agent systems and LLM orchestration

Soft Skills & Communication

• Excellent English communication skills (written and verbal)
• Ability to explain complex technical concepts to both technical and non‑technical audiences
• Strong problem‑solving and analytical thinking capabilities
• Self‑motivated with ability to work independently and drive projects to completion
• Collaborative team player who thrives in fast‑paced environments
• Passion for staying current with rapidly evolving LLM technologies
• Ability to balance research experimentation with production reliability requirements

NOTE: All personal data is collected for recruitment purposes only.