AI Engineer / AI工程師

Description

SWAG is building AI-native product capabilities powered by LLM-based agents and multi-modal intelligence. These systems are not experimental prototypes — they operate in production, interact with real users, and directly impact product experiences and business outcomes.

We are looking for a AI Engineer to design, build, and operate AI agent systems that integrate large language models, vision models, internal data platforms, and external tools. You will work closely with Data, Product, and Engineering teams to deliver scalable, reliable, and observable AI systems.

Responsibilities 

1. Design and implement LLM-based AI agents that support multi-step reasoning, tool / function calling, and contextual decision-making.

2. Build and maintain agent workflows that interact with:

• Internal data platforms and services
• External APIs and third-party tools
• Product events and user signals

3. Integrate computer vision and multi-modal models (vision + language) into agent workflows and product features.

4. Apply vision models for use cases such as content understanding, classification, moderation, or personalization.

5. Own the end-to-end lifecycle of AI systems, including:

• Prompt and agent versioning
• Model selection, upgrades, and deprecation
• Online inference services and deployment

6. Implement LLMOps practices, including:

• Automated evaluation of LLM outputs
• Monitoring quality, latency, usage, and cost
• Identifying and mitigating real-world failure cases

7. Collaborate with Data Engineers to integrate AI systems into existing infrastructure, CI/CD pipelines, and observability platforms.

8. Work closely with Data Scientists, Product Managers, and Software Engineers to translate business requirements into production-ready AI solutions.

Requirements

1. 5+ years of experience building and operating production AI or ML systems.

2. Hands-on experience with LLM-based systems or AI agents, including:

• Prompt engineering and prompt versioning
• Tool / function calling
• Multi-step, stateful, or context-aware workflows

3. Strong Python skills and solid software engineering fundamentals.

4. Experience operating AI systems with real users, including awareness of:

• Latency, throughput, and cost trade-offs
• Failure modes, fallback strategies, and guardrails
• Quality evaluation beyond traditional accuracy metrics

5. Familiarity with LLMOps / MLOps practices, such as:

• Evaluation frameworks for LLM outputs
• Monitoring and observability for AI systems
• CI/CD for models, prompts, or agents

6. Solid understanding of machine learning fundamentals (training, evaluation, inference).

7. Experience integrating:

• Vision or multi-modal models as components within larger systems
• Embedding models, vector databases, or retrieval systems

8. Experience deploying AI services using Docker, Kubernetes, and major cloud platforms (AWS, GCP, or Azure).

Good to have

1. Experience designing or contributing to shared AI platforms or internal frameworks.

2. Experience acting as a technical lead or mentoring other engineers.

3. Hands-on experience training or fine-tuning vision or multi-modal models.

4. Experience with RAG systems, vector databases, or knowledge-augmented agents.

5. Experience operating high-scale or real-time AI systems.

3 Rounds

1. HR Phone Interview

2. Take-Home Assignment (90mins)

3. Onsite Interview